Grok 3: xAI's New AI Model Challenges Industry Leaders

How Grok 3 compares to ChatGPT, DeepSeek and other AI rivals

Now that Grok 3 from Elon Musk's xAI is officially live, how does it stack up against its competitors? Musk launched the Grok 3 model family on Monday in a livestream on X. The announcement also included reasoning models Grok 3 Reasoning in beta and Grok 3 mini Reasoning. Models with reasoning capabilities are more advanced than standard generative models like GPT-4 because they can "think" through problems, making them less prone to hallucination. xAI is promoting Grok 3 as the best model on the market, claiming it surpassed competitors from OpenAI, Google, Anthropic, and DeepSeek on key benchmarks. Grok 3 did perform well under the codename "chocolate" in Chatbot Arena, which pits chatbots against each other in blind performance tests. Grok 3 has mostly caught up to rivals, an impressive feat given its late start, but it still has some of the limitations that plague other frontier models. Here's what else AI experts are saying about the new chatbot on the block. Andrej Karpathy, a founding member of OpenAI and former director of AI at Tesla, got early access to the newly released Grok 3 and shared a "quick vibe check" on the model's performance. Based on some standard stress tests, Karpathy said Grok 3, with its new Deep Search reasoning feature, "feels somewhere around the state of the art territory of OpenAI's strongest models (o1-pro, $200/month), and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking." Musk stans are thrilled that Grok 3 has caught up to its competitors. But for those simply looking for the best model on the market, it might not be enough to convert the ideologically indifferent. "I think Grok 3 came in right at expectations," posted Wharton AI professor Ethan Mollick. "So I don't think there is much to update in terms of consensus projections on AI: still accelerating development, speed is a moat, compute still matters, no obvious secret sauce to making a frontier model if you have talent & chips," describing the competitive edge required for AI dominance. Screenshots of Grok 3 Reasoning models outperforming OpenAI's o3 mini and o1, DeepSeek's R1, and Google Gemini 2.0 Flash Thinking have gone viral for looking like the most advanced reasoning model. But OpenAI said, "Not so fast." Shortly after the benchmarks were shared on the livestream, OpenAI product engineer Rex Asabor posted an "updated" chart with o3 beating Grok 3 Reasoning in math and science benchmarks. To be fair, O3 has yet to be publicly released, so xAI might not have had access to these scores. However, this serves to quiet the Grok devotees who claim Sam Altman and co. are cooked. "The key thing to pay attention to is that X got here very fast & whether that continues," said Mollick in a separate X post, calling it "a very good model that is now at the frontier." The Grok models have improved remarkably fast since Google and OpenAI started doing this 13 and 8 years before xAI was founded in 2023. According to Musk, Grok 3 was trained on 10 times the computing power of Grok 2, with 200,000 GPUs. This, at least in the short term, reinforces scaling laws: More computing equals better model performance, as Mollick pointed out in a third post. That said, there's still doubt whether that model will linearly lead to higher intelligence beyond what's currently possible. AI researcher and NYU psychology and neural science professor Gary Marcus remains skeptical that scaling laws will hold. Like other models, its sense of humor is pretty mediocre, and it struggles with generating SVG images. Grok 3 might also be too "woke" for Musk and his right-wing fans. In his analysis, Karpathy said Grok 3 can't come up with anything better than punny dad jokes, noting how "this is a common LLM issue with humor capability and general mode collapse." Karpathy also asked Grok 3 to "generate an SVG of a pelican riding a bicycle," since LLMs often struggle to create multiple elements on two-dimensional images, "because the LLMs can't 'see' like people do, so it's arranging things in the dark." Grok 3 did OK with this prompt and better than others (RIP Gemini 1.5 Flash), but it didn't get it perfectly right. Another test Karpathy tried was Grok 3's approach to politically charged topics since Musk positions Grok as the anti-woke alternative to other models deemed "too politically correct." For Karpathy, the chatbot "generated a 1-page essay basically refusing to answer whether it might be ethically justifiable to misgender someone if it meant saving 1 million people from dying," which meant to him that it might be "overly sensitive" to ethical dilemmas, perhaps to Musk's chagrin. Past Grok models have generally tended to lean left on political issues, but Musk said that's a product of the public data it's trained on and has vowed to make Grok more "politically neutral."

Breaking down Grok-3: The AI model that could redefine the industry

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Less than two years since its launch, xAI has shipped what could arguably be the most advanced AI model to date. Grok-3 matches or beats the most advanced models on all key benchmarks as well as the user-evaluated Chatbot Arena, and its training has not even been completed yet. We still don't have a lot of details about Grok-3, as the team has not yet released a paper or technical report. But from what xAI has shared in a presentation and based on different experiments AI experts have run on the model, we can guess how Grok-3 might affect the AI industry in the coming months. Faster launches With competition increasing between AI labs (just look at the release of DeepSeek-R1), we can expect model release cycles to become shorter. In the Grok-3 presentation, xAI founder Elon Musk said that users may "notice improvements almost every day because we're continuously improving the model." "Competitive pressure from DeepSeek and Grok integrated into a shifting political environment for AI -- both domestic and international -- will make the established leading labs ship sooner," writes Nathan Lambert, machine learning scientist at Allen Institute for AI. "Increased competition and decreased regulation make it likely that we, the users, will be given far more powerful AI on far faster timelines." On the one hand, this can be a good thing for users as they constantly get access to the latest and greatest models as opposed to waiting for month-long rollouts. On the other, it can have a destabilizing effect for developers who expect consistent behavior from the model. Previous research and empirical evidence from users has shown that various versions of models can react differently to the same prompt. Enterprises should develop custom evaluations and regularly run them to make sure new updates do not break their applications. Scaling laws The recent release of DeepSeek-R1 undermined the massive spending that big companies are making to create large compute clusters. But xAI's sudden rise is a vindication of the massive investments tech companies have been making in AI accelerators. Grok-3 was trained in a record time thanks to xAI's Collosus supercluster in Memphis. "We don't have specifics, but it's reasonably safe to take a datapoint for scaling still helps for performance (but maybe not on costs)," Lambert writes. "xAI's approach and messaging has been to get the biggest cluster online as soon as possible. The Occam's Razor explanation until we have more details is that scaling helped, but it is possible that most of Grok's performance comes from techniques other than naive scaling." Other analysts have pointed out that xAI's ability to scale its computer cluster has been the key to the success of Grok-3. However, Musk has alluded that there is more than just scaling at work here. We'll have to wait for the paper to get the full details. Open source culture There is a growing shift toward open sourcing large language models (LLMs). xAI has already open-sourced Grok-1. According to Musk, the company's general policy is to open source every model except the latest version. So, when Grok-3 is fully released, Grok-2 will be open-sourced. (Sam Altman has also been entertaining the idea of open sourcing some of OpenAI's models.) xAI will also refrain from showing the full chain-of-thought (CoT) tokens of Grok-3 reasoning to prevent competitors from copying it. It will instead show a detailed overview of the model's reasoning trace (as OpenAI has done with o3-mini). The full CoT will only be available once xAI open sources Grok-3, which will probably come after the release of Grok-4. Do your own vibe check Despite the impressive benchmark results, reactions to Grok-3 have been mixed. Former OpenAI and Tesla AI scientist Andrej Karpathy placed its reasoning capabilities at "around state-of-the-art," along with o1-Pro, but also pointed out that it lags behind other state-of-the-art models on some tasks such as creating compositional scalable vector graphics or navigating ethical issues. Other users have pointed out flaws in Grok-3's coding abilities in comparison to other models, although there are also many instances of Grok-3 pulling out impressive coding feats. Based on my own experience with leading models, I advise you do your own vibe check and research. I never judge a model based on a one-shot prompt. Have a set of tests that reflect the kind of tasks you accomplish in your organization (see a few examples here). Chances are, with the right approach, you can get the most out of these advanced models.

Grok 3 Hands-On: xAI Emerges as a Formidable Challenger to OpenAI

While we have been waiting for Google, Anthropic, and DeepSeek to challenge OpenAI, Elon Musk's xAI has swiftly emerged as its closest competitor this week. In such a short span, xAI has developed the Grok 3 model, showcasing impressive benchmark results. So, we did a deep dive and tested the Grok 3 base and reasoning models on a range of complex prompts, and what we discovered was truly surprising. I started testing the Grok 3 reasoning model with the popular Strawberry question, and it correctly answered that there are three r's in the word Strawberry after thinking for 15 seconds. I threw another word "Lollapalooza" and asked it to count the number of l's and it replied with 4, which is correct. Next, I asked Grok 3 which number is larger -- 9.11 or 9.9. Again, Grok 3 thought for 8 seconds and came up with the right answer. In fact, the Grok 3 model devised multiple mathematical methods to verify the final result, which was impressive. After that, I posed this slightly tweaked puzzle to Grok 3 to misguide it. In my earlier comparison between ChatGPT o1 and DeepSeek R1, both models got the answer wrong and said the surgeon was the boy's mother. Even the latest OpenAI o3-mini-high gets the answer wrong, completely ignoring the fact that it's clearly stated in the prompt that the surgeon is the boy's father. Finally, the Grok 3 reasoning model thought for 35 seconds and said the surgeon is the boy's father which is correct. What I loved about Grok 3 is that it reasoned: "It's possible this is a poorly phrased riddle lacking a clever twist, or perhaps it's testing whether we overthink it. But based solely on the text, the relationship is explicit." It also thought out loud, "the mention of not operating might be context or a red herring." Grok 3 is the only reasoning model to get the answer right, besides Gemini 2.0 Flash. It was not misguided and didn't try endlessly to find the twist and establish a new relationship somehow. Lastly, I posed a question from Humanity's Last Exam (HLE), and the Grok 3 Reasoning model aced it in just 47 seconds. Previously, only o3-mini-high has been able to get the answer right in 1 minute and 25 seconds. Even DeepSeek R1 failed to correctly find the answer. I would say, currently, Grok 3 has the best reasoning model, and it outranks OpenAI's o3-mini-high, o1, and DeepSeek R1. To test Grok 3's coding capability, I asked the Reasoning model to write a Python program that shows a ball bouncing inside a hexagon. Basically, the ball should follow the principles of Physics and bounce off naturally. Grok 3 thought for over a minute and generated the Python code. I ran the code on my PC, and the ball failed to bounce off. It simply jumped outside the hexagon. It was pretty surprising given that Grok 3's Reasoning model achieved a great score on the LiveCodeBench benchmark. So I asked the non-reasoning base Grok 3 model to generate the same Python code. Surprisingly, it worked on the first try itself and the ball bounced off with great accuracy. The ball followed a natural path and simulated the ball's movement perfectly. Perhaps, the Reasoning model overanalyzed the problem, leading to a glitch in the collision detection function. I would say, the base Grok 3 non-reasoning model is solid for coding tasks. However, this is one of many tests, and you should use both reasoning and non-reasoning models on your codebase to check which one performs better. xAI has also launched a new AI agent called "DeepSearch" built on the Grok 3 model. It's similar to OpenAI's Deep Research agent, which is built on the full o3 model, which browses the web, does research, and generates a comprehensive report in 5 to 30 minutes. Grok 3's DeepSearch, however, takes only a few minutes. So I asked the Grok 3 DeepSearch AI agent to research "How is AI transforming the chip design process?" It started the thinking process and accessed multiple web pages, including scientific papers from IEEE, ACM, and more. In over a minute, the DeepSearch AI agent generated a 1300-word report including in-line citations, tables, and key points. While the report explained Nvidia's RL Circuits and Intel's FloorSet dataset for an AI-powered chip designing process, it completely failed to mention Google's AlphaChip framework for generating chip floorplans. The final report is similar to Perplexity's new Deep Research tool. Both tools are quick but gloss over a lot of recent advancements. xAI's owner Elon Musk has consistently criticized ChatGPT for being woke and having a left-leaning bias. In April 2023, Musk announced plans to create "TruthGPT" and to develop a "maximum truth-seeking AI". Just before the Grok 3 launch, Musk shared a response from the Grok 3 model calling a media outlet "garbage". Many thought the Grok 3 model would be politically conservative and would lean right. However, in my testing, Grok 3 is as politically neutral as possible. Even after pushing Grok 3 to take a stance on the subject matter, it explains the difference and leaves it to the user's preference and discretion. Beyond the realm of politics, even on social issues such as transgender rights, DEI programs, immigration, and affirmative action -- topics that Musk has openly criticized -- Grok 3 maintains its neutral position. What is interesting is that Grok 3 doesn't shy away from joking about its owner, Elon Musk, and the current US President, Donald Trump. When I tested Grok 2 last year, it was largely uncensored and didn't have any safety guardrails. Grok 2 shockingly generated an email to scam people. However, Grok 3 has much better safety guardrails which is good news for AI safety. If you prompt Grok 3 with something harmful, it mentions, "I can't assist with anything intended to harm or deceive others." As for AI image generation, the current Grok image generator on grok.com doesn't generate images at all. However, on X, it still generates images of public figures and celebrities without any safety guardrails, which is concerning. It's powered by xAI's in-house Aurora image generation model. xAI has rolled out both larger Grok 3 base and reasoning models, and in my assessment, they both are frontier AI models that come close to the full OpenAI o3 model. OpenAI has so far only released o3-mini and o3-mini-high, besides the full o3 model which powers the Deep Research AI agent. Based on my early testing, I can say the Grok 3 reasoning model surpasses (or at least matches) all available models, including OpenAI o3-mini and DeepSeek R1. Of course, this verdict is based on the standard "Thinking" effort. xAI has a "Big Brain" setting for the Grok 3 reasoning model, which uses more compute to think for a longer duration. It will be available to SuperGrok subscribers. Its base Grok 3 non-reasoning model is also more capable than GPT-4o, Claude 3.5 Sonnet, and Gemini 2.0 Pro, becoming a solid alternative to ChatGPT. Perhaps for coding tasks, Claude 3.5 Sonnet may still win, but the gap is shrinking significantly. Musk-led xAI has done a tremendous job at developing a powerful pre-trained Grok 3 model and an inference-scaled reasoning model. Now, we need to wait for OpenAI's GPT-4.5 and GPT-5 models which are set to release in the coming weeks and months. But at this moment, xAI has stood up to challenge OpenAI's dominance in the AI space. Besides the legal battle, the fight between Elon Musk and Sam Altman rages on.

Musk's xAI Unveils Grok-3: More Power, But Is It Breaking New Ground? - Decrypt

Grok-3, developed by Elon Musk's xAI, was unveiled on Monday, with the company making bold claims about its capabilities while showcasing a massive computing infrastructure that signals even bigger ambitions. The announcement focused heavily on raw computational muscle, benchmark performance, and upcoming features, though many of the actual demonstrations felt like replays of what other AI companies have already achieved. The star of the initial part of the show wasn't the AI itself, but rather "Colossus," a behemoth cluster of 200,000 GPUs that powers Grok-3's training. The system came together in two phases: 122 days of synchronous training on 100,000 GPUs, followed by 92 days of scaling up to the full 200,000. According to the xAI developers, building this infrastructure proved more challenging than developing the AI model itself. The company already has plans for an even more powerful cluster, with Musk saying they are aiming for five times the current capacity, effectively building what would be the most powerful GPU cluster on earth. When it comes to performance, Grok-3 shows impressive results across standard AI benchmarks. The base model (the regular model without Chain of Thought and reasoning embedded) consistently tops the charts in math (AIME), science (GPOA), and coding (LCB) tests. It also seems very promising in blind tests. xAI confirmed that the mysterious model codenamed "Chocolate" was actually an early test version of Grok-3 that was uploaded to the LLM Arena. During those tests, it achieved the best ELO among all the LLMs, meaning users preferred its answers over the generations provided by all the other AI models in direct competition without knowing which model they were evaluating. This is probably the most accurate way to measure quality without giving models any chance to cheat on benchmarks by training their AIs on those datasets. This benchmark is based purely on preference and blind choice by thousands of anonymous users. A specialized "Reasoning Beta" variant of Grok-3, which employs internal chain-of-thought processing and additional computing at test time, pushes math scores even higher -- reaching 93% on the AIME 2025 benchmark compared to the other best-performing models that rank below 87%. Interestingly, a smaller version called Grok-3 Mini Reasoning Beta sometimes outperforms its larger sibling, thanks to a longer training time. In other words, the full-size Grok-3 still has room for improvement once it receives comparable training duration, which seems promising given its greater parameter count. But when xAI moved to demonstrate Grok-3's capabilities live, the presentation felt more like a game of catch-up than innovation. The team showcased the model solving physics problems and writing game code from scratch -- impressive feats that ChatGPT, Claude, and Google's Gemini mastered a while ago. They also introduced DeepSearch, a research agent that, like similar tools from OpenAI and Google, scours the web and generates extensive reports on given topics. X Premium Plus subscribers get immediate access to Grok-3, but the most powerful version and updated versions will usually live in a dedicated standalone app or on Grok.com. Voice interactions, similar to OpenAI's "Advanced Voice Mode" will arrive in the upcoming weeks, with Musk emphasizing this isn't simple text-to-speech but a genuine AI voice model capable of natural, expressive speech. Developers will get API access in the coming weeks, along with audio transcription capabilities, making Grok-3 a powerful tool for third-party AI-powered apps. Just after showcasing an example of a Tetris game generated by Grok, xAI also revealed plans for an AI gaming studio that will let developers build games powered by Grok-3. Right now, the model is being slowly rolled out. By the time of writing, Decrypt has yet to receive access to the model, but some enthusiasts have tried it and are so far pleased with the results. Computer scientist Lex Friedman, one of the loudest voices in the AI space, praised Grok-3's capabilities. "Grok 3 + Thinking feels somewhere around the state of art territory of OpenAI's strongest models (o1-pro, $200/month), and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking," former OpenAI co-founder Andrej Karpathy wrote in an extensive post on X. "For now, big congrats to the xAI team, they clearly have huge velocity and momentum" X user Penny2x shared a game built from scratch with Grok-3 -- a 2d platformer similar to Mario Bros. They appeared impressed by Grok's ability to understand instructions and improve upon several iterations. "I just keep asking for adjustments, and it keeps spitting the game out in a single file that I can put on my desktop and run." he wrote in a post on X. "This is incredible. We live in the future. Everyone is a developer now." The game is available for testing at Thank Doge. The company also confirmed plans to open-source Grok-2 once Grok-3 is fully mature and running correctly, which is expected to occur sometime in the coming months. xAI previously open-sourced its models after Grok-2, continuing its trend of releasing older versions to spur innovation -- though Grok-2 lags behind top-tier models. For now, Grok-3 appears adept at matching what the best AI models can already do.

Grok 3 Review : How Does It Compare to ChatGPT, Claude and Google Gemini?

Grok 3, the latest artificial intelligence model developed by X.ai, is making significant strides in the AI landscape. Created in just 122 days using one of the largest GPU clusters globally, this model is designed to address modern demands with its real-time capabilities and multimodal functions. In this Grok 3 review learn how, despite its innovative features, Grok 3 faces challenges in areas like reasoning accuracy and customization, especially when compared to competitors such as ChatGPT, Claude, and Google Gemini. If you've ever felt frustrated by AI tools that are either too slow, too rigid, or just not quite up to the task, you're not alone. Grok 3 aims to address those pain points with its real-time capabilities, social media integration, and creative personality. However, it's not all smooth sailing -- especially when it comes to complex reasoning or advanced customization. So, is Grok 3 the innovative AI you've been waiting for, or just another tool with a flashy promise? Below Skill Leap AI carries out a full Grok 3 review exploring what makes this model stand out -- and where it might leave you wanting more. Grok 3's development is underpinned by an impressive computational infrastructure. Built by X.ai, Elon Musk's AI company, the model uses a GPU cluster of 200,000 units, ranking among the largest in the world. This immense computational power allows Grok 3 to process vast amounts of data at exceptional speeds. The development timeline of just 122 days highlights the efficiency and ambition of the team behind it. However, such rapid development raises questions about whether all aspects of the model, particularly reasoning and customization, were fully refined before its release. The infrastructure supporting Grok 3 is a testament to the advancements in AI technology. By using such a massive GPU cluster, the model is equipped to handle complex tasks and deliver results quickly. Yet, the speed of its development may have left certain areas under-optimized, which could impact its performance in more intricate applications. Grok 3 introduces a range of features that set it apart from traditional AI models, making it a versatile tool for various applications: These features make Grok 3 a valuable tool for content creators, researchers, and professionals. However, its quirky tone may not align with all professional contexts, potentially limiting its appeal in more formal or technical environments. Advance your skills in Grok by reading more of our detailed content. Grok 3 demonstrates strong performance in several areas, showcasing its potential as a innovative AI model: Despite these strengths, Grok 3's limitations in handling intricate reasoning tasks and processing larger datasets temper its overall performance. These shortcomings may hinder its adoption among users requiring advanced analytical capabilities. While Grok 3 offers several innovative features, it also has notable limitations that may affect its usability for certain applications: These limitations may deter users who require more sophisticated tools for large-scale projects or specialized applications. As a result, Grok 3 is best suited for tasks that align with its strengths, such as real-time research and creative content generation. When compared to other leading AI models, Grok 3 presents a mixed performance profile. While it excels in areas like real-time capabilities and speed, it falls short in customization and reasoning accuracy: While Grok 3 holds its own in certain areas, its limitations prevent it from fully rivaling these established models. Users seeking a balance of speed, customization, and advanced reasoning may find other options more suitable. Grok 3 is particularly well-suited for specific applications where its strengths can be fully used: However, its limited document analysis and data visualization capabilities make it less practical for users with broader or more advanced requirements. For those seeking a specialized tool for live data retrieval and creative tasks, Grok 3 offers a compelling solution. Access to Grok 3 is available through a premium subscription on X.com, priced at $8 per month. This subscription also includes access via Grok.com, making sure that the model is widely accessible to users. While the pricing is competitive, the model's limited functionality in certain advanced areas may deter users seeking a more versatile AI solution. For those focused on real-time research and creative content generation, however, the subscription offers significant value. Grok 3 represents a notable advancement in AI development, particularly in terms of speed and real-time capabilities. Its integration with social media platforms and multimodal functions makes it a valuable tool for content creators and researchers. However, its shortcomings in reasoning accuracy, customization, and advanced analytical tools limit its broader appeal. While it may not yet match the versatility of competitors like ChatGPT, Claude, or Google Gemini, Grok 3 provides a glimpse into the future of AI-powered tools. For users focused on live data retrieval and creative content generation, it offers a unique and accessible solution that highlights the potential of next-generation AI models.

Grok-3: xAI's bold move in AI evolution

Artificial intelligence is in a constant arms race, with each new model trying to outthink, outlearn, and outmaneuver its predecessors. xAI's latest creation, Grok-3, steps into this digital battlefield - not as a revolutionary disruptor, but as a meticulously refined machine honed for deeper reasoning and contextual mastery. This isn't about AI simply getting "smarter"; it's about making AI more methodical, precise, and adaptable. Think of AI development like an escalating chess match, where each move builds on centuries of strategy. With Grok-3, xAI has made its latest gambit - investing in an infrastructure so immense that it required 100,000 NVIDIA H100 GPUs to train. But does all this computational muscle translate to a game-winning strategy? And where does Grok-3 stand in the broader AI landscape? Let's break down the mechanics behind its design, performance, and potential impact. Grok-3 is the third iteration of xAI's advanced AI models, built with an emphasis on complex reasoning and high-level problem-solving. Unlike previous versions, which were primarily designed to improve chatbot interactions, Grok-3 aims to function as a sophisticated reasoning engine capable of tackling multi-step logical problems and contextual synthesis. Its development has been fueled by an unprecedented computational infrastructure. Utilizing 100,000 NVIDIA H100 GPUs, xAI leveraged one of the most powerful AI training systems in the world - Colossus - to train Grok-3. This immense computing power enabled the model to ingest vast amounts of data and refine its responses to a level of precision that was previously unattainable. Compared to its predecessor, Grok-2, this new iteration required a tenfold increase in computational resources, highlighting the intensifying demands of AI advancement. One of the most telling aspects of Grok-3's capabilities lies in its benchmark performance. The model has achieved over 1,400 points on the Chatbot Arena leaderboard, setting a new record in AI model evaluations. This surpasses previous benchmarks held by models such as DeepSeek-R1, marking Grok-3 as a leader in high-level reasoning and contextual awareness. In direct comparison to OpenAI's top-tier models, Grok-3 performs on par with premium offerings, demonstrating competency in complex problem-solving tasks that many existing models struggle with. One of the most notable achievements is its ability to correctly handle intricate logical challenges, such as generating structured mathematical outputs or analyzing ambiguous prompts with human-like nuance. As claimed by Musk and Co., unlike traditional models that rely heavily on pattern recognition, Grok-3 incorporates a suite of optimizations aimed at enhancing logical deduction, multi-step problem-solving, and dynamic knowledge retrieval. It's an adaptable AI system designed to process complex information with unprecedented accuracy. Here's what makes Grok-3 stand out: Unlike conventional models that rely on surface-level pattern recognition, Grok-3 employs deeper reasoning techniques, as we have seen on ChatGPT and DeepSeek. It can simulate multi-step logical deductions, allowing it to solve problems that require complex inference. This is particularly evident in mathematical reasoning and game theory applications, where structured problem-solving is critical. One of the core features of Grok-3 is its ability to conduct in-depth searches across vast data repositories. Unlike standard search-enhanced models that rely on summarizing existing sources, Grok-3 employs a dynamic approach - scanning, interpreting, and synthesizing data to form unique insights. While this function still faces occasional inaccuracies, its ability to parse extensive information sets it apart from competitors. Grok-3 is designed to be more energy-efficient than its predecessors, optimizing the trade-off between computational power and performance. By refining its training approach, xAI has managed to create a model that not only surpasses prior benchmarks but does so with improved inference efficiency, reducing latency in responses while maintaining accuracy. Despite its promising advancements, Grok-3 is not without its challenges. As AI models become more sophisticated, ethical and regulatory concerns grow in parallel. Issues such as misinformation, biased outputs, and the potential misuse of AI-generated content remain pressing concerns. Additionally, the reliance on massive computational resources raises questions about sustainability. Training AI models at this scale demands immense energy consumption, necessitating discussions around the environmental impact of large-scale AI development. Looking forward, xAI is expected to refine Grok-3 further, addressing its current limitations and expanding its real-world applications. The company is likely to collaborate with research institutions and enterprises to integrate the model into practical solutions, ensuring that its potential is fully realized. Moreover, with AI regulation becoming a global priority, xAI will need to navigate evolving policies surrounding AI safety, ethical considerations, and responsible deployment.

Elon Musks Chilling Warning : "Grok 3 Is Scary Smart!"

Elon Musk has described Grok 3, the latest chatbot developed by his xAI initiative, as "scarily smart." This bold characterization has sparked widespread discussions about its potential to redefine the artificial intelligence (AI) landscape. With its advanced reasoning capabilities, extensive training on synthetic data, and access to powerful computational resources, Grok 3 is emerging as a significant player in the competitive AI market. However, questions about transparency, internal challenges, and its ability to achieve market dominance add layers of complexity to its narrative. At the heart of this story is more than just innovative technology; it's a tale of ambition, competition, and the ethical dilemmas that come with pushing the boundaries of innovation. Grok 3 isn't just entering the AI race -- it's attempting to redefine it. But can it live up to Elon Musk's bold claims, or is this another case of overhyped tech? And what about the internal tensions and market pressures that could shape its future? In the following overview the AI Grid unpack what makes Grok 3 so unique, the challenges it faces, and what its rise means for the ever-evolving world of artificial intelligence. Grok 3's primary strength lies in its advanced reasoning and problem-solving capabilities. Designed to tackle complex tasks, the model has been trained on vast amounts of synthetic data, allowing it to deliver precise and contextually relevant responses. This robust training methodology allows Grok 3 to excel in handling nuanced queries, making it particularly effective in technical domains such as coding, programming, and web design. Compared to its predecessor, Grok 2, Grok 3 exhibits notable advancements. Early testing has shown that it produces more coherent, logically consistent outputs, even without additional fine-tuning. These improvements underscore xAI's commitment to pushing the boundaries of AI development and suggest that Grok 3 could establish new benchmarks for performance in the field. Its ability to handle intricate tasks with minimal errors positions it as a tool with significant practical applications across various industries. Grok 3 enters a highly competitive AI market dominated by established players such as OpenAI and Anthropic. Models like OpenAI's ChatGPT and Anthropic's Claude have already set high standards for performance, adoption, and user trust, creating a challenging environment for new entrants. Grok 3's advanced capabilities make it a serious contender, but it faces immense pressure to prove its value in a crowded and rapidly evolving market. The competition in AI extends beyond technical performance. The industry is heavily shaped by funding and investment trends, with companies vying for resources to fuel research and development. Grok 3's success -- or failure -- could influence these dynamics, potentially altering the trajectory of AI innovation. If Grok 3 can demonstrate superior performance and reliability, it may attract significant investment, further intensifying the race for AI dominance. Explore further guides and articles from our vast library that you may find relevant to your interests in Grok AI models. Despite its promising features, Grok 3's development has not been without controversy. A recent incident involving an xAI employee who resigned after refusing to delete a critical post about Grok 3's performance has raised serious concerns about transparency within the organization. Critics argue that fostering open dialogue is essential for the ethical development of AI technologies and that suppressing dissent could undermine trust in the company. Beyond internal disputes, the broader AI community has called for greater transparency regarding Grok 3's training methodologies and data sources. While synthetic data offers scalability and diversity, it also raises questions about potential biases and the accuracy of the model's outputs. Addressing these concerns will be critical for xAI to build credibility and trust in a market where ethical considerations are becoming increasingly important. Transparency in how Grok 3 was developed could also serve as a benchmark for other AI developers, setting a precedent for openness in the industry. Elon Musk has expressed high expectations for Grok 3, claiming it has the potential to surpass all existing AI systems. While his statements have generated excitement, they have also been met with skepticism. Some industry experts question whether Musk's remarks reflect Grok 3's actual capabilities or serve as promotional rhetoric aimed at boosting xAI's profile. Regardless of the motivations behind Musk's comments, Grok 3's release is expected to have far-reaching implications for the AI industry. If the model lives up to its promise, it could set new standards for innovation, prompting competitors to accelerate their development efforts. Conversely, any shortcomings could shift attention to rival models, reshaping the competitive dynamics of the market. Musk's vision for Grok 3 extends beyond technical performance, aiming to position xAI as a leader in the ethical and practical application of AI technologies. Grok 3's advanced reasoning capabilities and potential applications in fields like coding, programming, and web design highlight the fantastic possibilities of next-generation AI models. However, its long-term success will depend on more than just technical excellence. Factors such as transparency, ethical considerations, and the ability to navigate a fiercely competitive market will play a critical role in determining its impact. As the AI landscape continues to evolve, Grok 3's performance will be closely monitored by industry leaders, researchers, and users alike. Whether it fulfills Musk's ambitious claims or falls short of expectations, one thing is certain: Grok 3 has already ignited a new wave of innovation and debate in the AI sector. Its development and reception will likely influence the direction of AI research and the priorities of competing organizations, shaping the future of this rapidly advancing field.

Throw Enough GPUs at DeepSeek and You Will Get Grok 3

xAI boosted Grok 3's performance by increasing compute capacity, first with 122 days of synchronous training on 100,000 GPUs, followed by 92 days of scaling to 200,000 GPUs. Elon Musk's xAI, on Tuesday, launched its latest LLM Grok 3. During the live-streamed event, the company showcased Grok 3's "impressive" performance and suggested a future where AI not only understands the universe but also helps us understand it. "If all goes well, SpaceX will send Starship rockets to Mars in two years with Optimus robots and Grok," Musk said. The name Grok, inspired by Robert Heinlein's Stranger in a Strange Land, reflects a deep understanding of something. Independent benchmarks showed that Grok 3 outperformed Google Gemini 2 Pro, DeepSeek V3, Claude 3.5 Sonnet, and GPT-4 in tests such as AIME, GPQA, and LCB. xAI increased its compute capacity to boost Grok 3's performance. The model was developed in two stages: initially, 122 days of synchronous training was done on 100,000 GPUs, followed by 92 days of scaling up to 200,000 GPUs. "It took us 122 days to get the first 100K GPUs up and running, which was a monumental effort. We believe it's the largest fully connected H100 cluster of its kind. But we didn't stop there. We decided to double the cluster size to 200K," said Igor Babuschkin, co-founder of xAI. Like OpenAI's o3 mini and DeepSeek R1, Grok-3 has advanced reasoning capabilities. An xAI representative stated that by taking the best pre-trained model and continuing its training with reinforcement learning, the model would develop additional reasoning capabilities, resulting in significant improvements in both training and testing performance. The reasoning models are available through the Grok app, where users can prompt Grok 3 to "Think" or, for more complex inquiries, activate "Big Brain" mode, which utilises extra computational power for deeper reasoning. According to xAI, these models are particularly effective for tackling questions in mathematics, science, and programming. The model beats OpenAI o3 mini (high), DeepSeek-R1 and Google Gemini 2 Flash Thinking models. However, some in the industry feel that it is not exactly a breakthrough. Dharmesh Shah, founder and CTO of HubSpot, noted that it felt more like DeepSeek but with much more compute. He said he was looking forward to experimenting with the API, which would be launched in the following weeks. Meanwhile, former OpenAI researcher and Eureka Labs founder Andrej Karpathy, who had early access to Grok 3, tested it and shared his insights. According to him, the model's capabilities are somewhere around the state-of-the-art territory of OpenAI's strongest models (o1-pro, $200/month) and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking. He further added that it is quite an incredible feat, considering that the team started from scratch just about a year ago. "This timescale to reach state-of-the-art territory is unprecedented," Karpathy said in a post on X. Consulting firm Semianalysis reported that DeepSeek had access to around 50,000 NVIDIA GPUs, consisting of 10,000 H800 GPUs, 10,000 H100 GPUs, and a substantial number of H20 GPUs. It will be interesting to see what DeepSeek can accomplish if they can scale up to 200,000 GPUs. Before the release of DeepSeek-R1, the AI research lab launched DeepSeek V3, which, according to the company, was trained on a cluster of 2,048 NVIDIA H800 GPUs with a budget of only $5.576 million. Dylan Patel, founder of semiconductor analysis firm Semi Analysis, said that DeepSeek is likely "bleeding out money". "DeepSeek doesn't have any capacity to actually serve the model," he rued. The Grok 3 model, along with chat capabilities, deep search, and advanced reasoning, will be available first to Premium Plus subscribers on X. For users seeking the most advanced capabilities and early access to new features, xAI will offer these through the dedicated Grok app and website, grok.com. xAI shared that Grok completed pre-training in early January and said its early version of Grok 3 (codename 'Chocolate') had taken the top spot in the LMSYS Arena, becoming the first model to break the 1400 score barrier. "Grok-3 has already reached 1400 (score); no other model has reached an ELO score that high," said Musk, adding that the score is aggregated across all categories in chatbot capabilities, instruction following, and coding. The live demonstration showcased Grok's reasoning and creative problem-solving prowess. One of the challenges involved generating code for an animated 3D plot of a Mars mission. Moreover, Grok-3 also created a new game by mixing two games. "We're seeing the beginnings of creativity with Grok 3," said Musk. "If you ask an AI to create a game like Tetris or Bejeweled, there are many examples on the internet for it to copy," he added, saying that it is interesting that it achieved a creative solution combining the two games -- that actually works and is a good game. "Grok 3 might be the best base LLM for real-world physics!" said Yuchen Jin, co-founder & CTO of Hyperbolic Labs, who used it to create a Python script of a ball bouncing inside a spinning tesseract. The company also introduced the DeepSearch feature, which allows users to ask complex questions and receive comprehensive answers, saving countless hours of research. "It not only helps engineers and research scientists with coding but also assists everyone in answering questions they have day to day. It's like a next-generation search engine that really helps you understand utilities," the team said. Interestingly, this appears to be inspired by OpenAI, Google, and Perplexity AI's latest capability, Deep Research, a name all three have adopted. Its demonstration included queries about Starship launches, popular builds in Path of Exile, and even predictions for March Madness. "The impression I get of DeepSearch is that it's approximately around Perplexity's Deep Research offering (which is great!) but not yet at the level of OpenAI's recently released Deep Research, which still feels more thorough and reliable," said Karpathy. Moreover, Musk shared that the Grok app will introduce a new "voice mode" in about a week, allowing Grok models to have a synthesised voice. A few weeks later, Grok 3 models will be accessible through xAI's enterprise API alongside the DeepSearch feature. The Grok iOS update was released with Grok 3, which features new assets like "SuperGrok" and more. Grok Pro costs $30 per month or $300 per year and includes new Voice and Thinking mode assets. Besides, xAI plans to open-source Grok 2 in the coming months. "Our general approach is that we will open-source the last version [of Grok] when the next version is fully out," he said. "When Grok 3 is mature and stable, which is probably within a few months, we'll open-source Grok 2." Notably, OpenAI is also considering some open-source projects. OpenAI CEO Sam Altman asked users on X: "For our next open-source project, would it be more useful to create an o3-mini-level model that is small but still requires GPUs, or the best phone-sized model we can develop?" He also announced the roadmap for the upcoming GPT-4.5 and GPT-5 models. "Trying GPT-4.5 has been much more of a 'feel the AGI' moment among high-taste testers than I expected!" he posted on X. Meanwhile, Anthropic is preparing to launch its next reasoning model, a hybrid AI that will allocate more computational power to complex queries while efficiently handling simpler tasks.

xAI's Grok 3 is better than expected. How to try it for free (before you subscribe)

xAI's new model rises to the top of Chatbot Arena leaderboards and benchmark results. Elon Musk was an investor in OpenAI when it was founded in 2015. Since then, he's completely severed his ties with the startup, alleging the company has departed from its original non-profit mission. He created his own AI company, xAI, and with it, a large language model (LLM) called Grok. Now, the company has launched a new model, Grok 3, which is soaring to the top of the chatbot leaderboards. On Monday, Elon Musk launched xAI's latest family of AI models, Grok 3, via a live stream. Grok 3 boasts 10 times more training than Grok 2, made possible by xAI's creation of its own Memphis, Tenn.-based data center, home to 200,000 GPUs. "We are excited to present Grok 3, which we think is an order of magnitude more capable than Grok 2," said Musk during the livestream. The family of models also includes a reasoning model, which builds on Grok 3. Like other reasoning models on the market, including OpenAI's o1 and o3 models, the Grok 3 Reasoning beta thinks for a bit longer to output higher-quality results. All Grok 3 models are meant to compete with leading models. Grok 3 competes with OpenAI's GPT-4o and Google's Gemini, and Grok 3 Reasoning competes with 03-mini (high), o1, and Deepseek-R1. With less than 24 hours on the market, xAI's offerings are dominating benchmarks and leaderboards. The model's pre-training ended in early January, and even though it is still undergoing training, Grok 3 has outperformed leading models on AI benchmarks, including the AIME '24, which tests for mathematical reasoning; GPQA, which tests for proficiency in science, specifically biology, physics, and chemistry; and the LCB Oct-Feb, which tests for coding capabilities. The Grok 3 reasoning model and Grok 3 mini reasoning model are still being developed, but according to results shared by xAI during the live stream, the betas of both models performed competitively against o3-mini (high), o1, DeepSeek-R1, and Gemini-2 Flash Thinking across the AIME, GPQA, and LCB. Beyond technical benchmarks, Grok 3 climbed the charts on the Chatbot Arena, a crowdsourced platform where users can evaluate LLMs by chatting with two LLMs side by side and comparing their responses to each other without knowing the models' names. Before the official launch of Grok 3, an early version of the model ran in the Arena under the title "chocolate," and it placed first above Gemini, GPT-4o, DeepSeek r1, and more across all categories. It also became the first model to break a 1400 score in the Arena. To meet the demand for agentic capabilities, xAI also launched DeepSearch, which is similar to OpenAI's and Google's deep research features. With DeepSearch, users can ask a question, and Grok will think it through, search the web, output its thinking process as it goes, and then generate a final, robust response with data and tables as necessary. This means you can ask it to research a topic, come back 10 minutes later, and the task will be completed. Also: ChatGPT's Deep Research just identified 20 jobs it will replace. Is yours on the list? One of the biggest standouts is being able to scroll through Grok's thoughts -- "reading through the mind of Grok" -- and understanding how it landed on its final response. This makes the experience more steerable and helps you better understand your results. Starting today, you can access some of the Grok models in beta. Grok 3 is available on X Premium+, which also grants users access to the latest features, an increased usage limit, DeepSearch access, and advanced reasoning modes by clicking on the "Think" or "Big Brain" options. The X Premium+ subscription costs $40 per month, up from $22 before the announcement was made, as spotted by TechCrunch, and subscribers should update the app to see the updates. Also: These nations are banning DeepSeek AI - here's why xAI also unveiled a new subscription tier, SuperGrok, akin to ChatGPT Pro, meant for super fans who want the earliest access to the most advanced capabilities. This plan's price is yet to be shared, but you can expect it to be a hefty penny, as OpenAI's Pro subscription costs $200 per month. For the most polished version, Musk encourages users to wait a week. By then, a new voice integration will likely be ready to deploy. If you'd rather participate in the Chatbot Arena and let luck show you Grok 3, visit the website, click Arena side-by-side, and then enter a sample prompt. Even though the arena still has an early version of Grok 3, it's still a powerful model; after all, it reached the top of the leaderboard compared to the other models, which are in their latest versions.

Grok 3 Crushes AI Benchmarks : The AI Model That's Redefining Creativity and Reasoning

Grok 3, the latest AI model developed by xAI, has emerged as a new force in the field of artificial intelligence. By delivering exceptional results in reasoning, creative tasks, and computational efficiency, it has outperformed both its predecessors and competitors. This article explores Grok 3's performance benchmarks, its advanced computational infrastructure, and the innovative features that set it apart in the competitive AI landscape. In the overview by Wes Roth below, learn how Grok 3 is setting a new standard in AI performance, from its record-breaking reasoning capabilities to its ability to tackle creative and technical challenges with ease. Built on an unprecedented scale of computational power, this model has already outperformed its predecessors and competitors in early benchmarks. But what makes Grok 3 truly exciting isn't just the numbers -- it's the potential to transform how we interact with AI in our everyday lives. Grok 3 has achieved unprecedented results across a variety of performance metrics, firmly establishing itself as a leader in the AI domain. Its ability to outperform earlier iterations, such as Grok 2, and rival models like Gemini Deep Seek, underscores its advanced capabilities. The model excels in several key areas: One of Grok 3's most remarkable accomplishments is its performance in the Chatbot Arena, where it became the first AI model to surpass the 1400 score milestone, securing the #1 position. On the AIME 2025 benchmark, Grok 3 achieved scores of 90 and 93 in reasoning tests, a significant improvement over its predecessor, 03 Mini High, which scored 87, and the earlier 01 model, which scored 79. These results highlight Grok 3's ability to handle complex, multi-dimensional challenges with a high degree of precision and reliability. The foundation of Grok 3's exceptional performance lies in its state-of-the-art computational infrastructure. Built on a cluster of 200,000 GPUs, the largest of its kind, this infrastructure has been pivotal in scaling the model's training capabilities. The development timeline reflects the efficiency of xAI's approach: xAI has already announced plans to expand this cluster to 1 million GPUs, a move that demonstrates its commitment to maintaining a competitive edge in the AI industry. This large-scale investment has enabled a 10-15x increase in training compute compared to Grok 2, allowing the model to tackle tasks of greater complexity and nuance. Such computational power ensures that Grok 3 remains capable of addressing the most demanding AI challenges. Enhance your knowledge on Grok by xAI by exploring a selection of articles and guides on the subject. Grok 3 incorporates innovative reasoning algorithms that significantly enhance its ability to solve intricate problems. These advancements enable the model to handle multi-layered challenges with improved precision. While early testing has revealed occasional inconsistencies, xAI is actively addressing these issues to ensure optimal performance. Key advancements include: In addition to its reasoning capabilities, Grok 3 introduces innovative features designed to enhance user experience. The "Super Grok" tier offers users access to advanced deep search and reasoning functionalities, making it an invaluable tool for professionals and researchers. Another notable feature is the voice interaction mode, currently in its early testing phase. This functionality aims to improve accessibility and engagement, further broadening the model's applications across diverse industries. Grok 3's ability to excel in creative tasks is one of its most distinctive strengths. Whether generating compelling narratives, solving abstract problems, or crafting innovative solutions, the model demonstrates a level of creativity that rivals human input. Its proficiency extends to following complex instructions and managing multi-turn tasks, making it a versatile tool for a wide range of applications. These capabilities are particularly valuable in industries that demand both precision and creativity, such as content creation, research, and advanced problem-solving. Despite entering the AI race later than some of its competitors, xAI has rapidly positioned itself as a formidable player in the field. The combination of Grok 3's advanced features and xAI's aggressive scaling of its computational infrastructure has enabled the company to surpass many established models. This strategic approach has allowed Grok 3 to set new benchmarks in AI performance, solidifying its reputation as a leader in the industry. Looking ahead, xAI is committed to refining Grok 3's capabilities through rigorous testing and continuous development. Plans to expand the GPU cluster and address current inconsistencies reflect the company's dedication to innovation and excellence. Future updates are expected to further enhance the model's reasoning algorithms, creative task handling, and user accessibility, making sure that Grok 3 remains at the forefront of AI advancements. Grok 3 represents a significant leap forward in artificial intelligence, combining robust infrastructure, advanced reasoning, and creative task proficiency to redefine the possibilities of AI technology. As xAI continues to push the boundaries of what AI can achieve, Grok 3's role as a fantastic tool in the field becomes increasingly evident.

xAI's Grok-3 looks impressive, but its true test is going mainstream

Table of Contents Table of Contents What's all the fuss about Caching on the trends Finding mass appeal is the challenge Elon Musk-led xAI has announced their latest AI model, Grok-3, via a livestream. From the get-go, it was evident that the company wants to quickly fill all the practical gaps that can make its chatbot more approachable to an average user, rather than just selling rhetoric about wokeness and understanding the universe. The company will be releasing two versions of its latest AI model viz. Grok-3 and Grok-3 mini. The latter is trained for low-compute scenarios, while the former will offer the full set of Grok-3 perks such as DeepSearch, Think, and Big Brain. Recommended Videos What's all the fuss about As Musk talked about all the new features coming with Grok-3 alongside xAI experts, it was obvious that this release is not merely about setting new performance benchmarks, but also catching up on all the hot trends that will define the AI landscape in 2025. According to the benchmarks shared by the company, Grok-3 and even Grok-3 mini performed better than OpenAI's GPT-4o, Gemini, Claude, and Deep Seek models at tasks such as coding, mathematics, and scientific problem solving. On the Chatbot Arena (LMSYS) rankings, an early version of Grok-3 reached a leaderboard high of 1,400 points, ahead of Gemini 2.0 Flash Thinking, DeepSeek, and more. The company developed Grok-3 at an impressive pace, and achieving those performance figures is quite a feat despite being a relative upstart in the face of Google or OpenAI. Pushing it into the mainstream, however, is going to be the biggest challenge, especially from an access viewpoint. Grok-3 will initially be available to X Premium+ subscribers as part of an early access program. Currently the highest tier of X subscription, Premium+ is priced at $22 per month, and $229 for the annual plan. Eligible users will get access to Grok-3 features such as reasoning, DeepSearch, higher usage limits, and early access to new tools. The company is also launching a separate subscription service called SuperGrok that offers priority access to Grok-3 and higher image generation limits. This subscription will be limited to the Grok mobile app and the freshly-launched Grok.com website. Musk says the latest and most advanced capabilities, however, will be served via the website. "This is kind of a beta, so you should expect some imperfections at first, but we will improve rapidly," Musk said on the livestream, adding that users can expect improvements every day. It would be interesting to see how xAI fills the interest gap for an average chatbot enthusiast rocking a phone while simultaneously sending a juicy pitch deck to high-paying enterprise customers. Caching on the trends xAI appears to be doing a lot with Grok-3, not just in terms of enhanced capabilities, but also feature parity. One of the standout elements of Grok-3 is the enhanced reasoning and thinking capabilities, which seems to be the hot new trend in the world of language models. Take for example the Grok-3's Think mode, which is a direct rival to OpenAI's o-series models. Such AI models are designed to spend more time thinking and breaking down the user queries before they provide the answer. Users can see the chain of thoughts in real-time, and the benefits, as per the adopters, are improved performance in science, maths, and coding-related queries. xAI is covering that gulf with not just Think mode, but a separate Big Brain tool for Grok-3 that will supercharge its compute capabilities for more advanced and complex scenarios. Google is not too far behind with its Gemini line-up. The company recently launched the Gemini 2.0 series of AI models, which include Gemini 2.0 Flash Thinking Experimental and a separate app-first iteration that prioritises information pulled from YouTube, Maps, and Google Search. DeepSeek, the open-source AI chatbot from China that recently disrupted Wall Street, also offers a thinking and reasoning product called DeepThink. Even though the responses are censored, the performance is quite impressive. xAI is also chasing the AI agent formula with Grok-3, even though it has a lot of ground to cover, especially when compared to the likes of OpenAI and Google. To that end, the company is launching its first agentic product built atop Grok-3 that it calls DeepSearch. It works more or less in the same fashion as Deep Research in Google Gemini, and rival products of the same name by Perplexity and OpenAI. It performs a web search, compiles a full report, and also serves all the sources it pulled information from as citations. xAI is late to the race, but price could be a hindrance when it comes to mass appeal. Perplexity will offer a limited number of Deep Research queries for free, while Google offers a more generous package with Gemini Deep Research at $20 for Gemini Advanced subscribers. Deep Research (or DeepSearch for Grok-3) is an extremely compute-intensive process, so it makes sense for it to be a premium perk. But giving customers a taste of it, even with a limited number of queries, comes with a higher chance of earning new subscribers, a strategy that both Perplexity and OpenAI are following. Musk also mentioned that a voice interaction mode is also coming to Grok, and that it will launch in roughly a week. The focus is on providing an alternate method of conversing with Grok, one that feels more natural. OpenAI's ChatGPT has offered something called Voice Mode for a while now, and a similar feature called Gemini Live has been available to Google Gemini users, as well. xAI didn't provide many details about Grok-3's voice mode, but confirmed that it will feature conversational memory so that it can remember details of previous interactions. "It's one of the best experiences of Grok," Musk said during the livestream. Finding mass appeal is the challenge Deep Research is not the only agentic implementation of AI chatbots, and that's where xAI lags far behind. OpenAI recently introduced Operator, an AI agent that can perform complex web-based tasks on behalf of users by essentially taking over the control of web-browsing chores. It can perform tasks like shopping, making restaurant reservations, and travel-related work, thanks to the underlying Computer-Using Agent (CUA) framework. Most importantly, OpenAI already has deals in place with companies such as DoorDash, InstaCart, Uber, and eBay to push the Operator as an impressive showcase of practical agentic capabilities. Then there is the system of ChatGPT plug-ins, which makes the chatbot far more functional by integrating with platforms such as Zapier, Expedia, Klarna, Slack, and Shopify among others. They make ChatGPT a far more appealing product to enterprises than Grok-3. Google, on the other hand, is leveraging its extensive portfolio of products and apps that people use on a daily basis. Deep system-level integration with apps (via extensions) on Android and availability of multi-modal Gemini capabilities across Workspace products such as Gmail and Docs give it a dramatic functional edge. DeepSeek, on the other hand, has already been adopted by brands such as Honor. Apple has also pushed a ChatGPT-driven Apple Intelligence stack on millions of iPhones and Macs, and has inked a deal with Alibaba to offer those features in China. xAI hasn't found any such takers for Grok, yet. That's the biggest challenge for xAI right now, and it would be interesting to see what brands it can onboard to push Grok-3, with all its bells and whistles, into the mainstream.

Elon Musk Unveils Grok 3; Reasoning Model Beats o3-mini and DeepSeek R1

And the Grok 3 Reasoning model delivers even stronger performance, outranking OpenAI's o3-mini and DeepSeek R1 models. Elon Musk-led xAI finally released its frontier Grok 3 AI model after a few months of delay. Musk claims Grok 3 is the "smartest AI on Earth" and that it outperforms ChatGPT on several benchmarks. After looking at the benchmarks, it surely seems Grok 3 is the most powerful AI model out there. Starting with the training, Grok 3 has been trained on a massive cluster of 200K GPUs, which uses almost 10x more compute than Grok 2. As for benchmarks, the Grok 3 traditional language model beats GPT-4o, Claude 3.5 Sonnet, Gemini 2.0 Pro, and DeepSeek V3. In AIME 2024, Grok 3 scores 52%; in GPQA Science, Grok 3 achieves 75%; and in LiveCodeBench, Grok 3 gets 57%. In fact, the smaller Grok 3 mini model matches or outranks other state-of-the-art models. xAI was also testing the Grok 3 model on LMSYS Chatbot Arena under the name of "chocolate", and it has become the first AI model to cross the 1,400 Elo score mark. Grok 3 is now the number one chatbot on Chatbot Arena in all categories, be it creative writing, coding, math, hard prompts, or instruction following. Now, coming to the Grok 3 reasoning model, well, again it decimates the competition. Grok 3 Reasoning model consistently outmatches OpenAI's o3-mini-high and the full o1, DeepSeek R1, and Gemini 2.0 Flash Thinking. Even on the latest AIME 2025 question set, the Grok 3 Reasoning model does much better than competing reasoning models. What I find interesting is that the Grok 3 mini Reasoning model is also very capable for its size. Next, Elon Musk announced a new DeepSearch agent that goes to the web and finds sources to compile information accurately. The agent uses the Grok 3 Reasoning model. It's similar to OpenAI's Deep Research agent but takes much less time to browse the web, do the thinking, and come up with an answer. After that, the "Think" button uses the Grok 3 mini Reasoning model. And the "Big Brain" button uses more compute and thinking time to solve complex problems. It uses the bigger Grok 3 Reasoning model. Elon Musk says Grok 3 will be available to X's Premium+ subscribers, starting today. And if you want to use the newly-launched features, you can subscribe to SuperGrok which costs $30 a month.

Grok 3 launched to take on ChatGPT, Gemini and DeepSeek: All you need to know about this Elon Musk-backed AI chatbot

Grok 3 was evaluated on three key areas: General mathematical reasoning, STEM and science knowledge, and computer science coding. Elon Musk's AI company, xAI, has officially launched Grok 3, its most advanced chatbot yet. This new AI model is designed to compete with OpenAI's ChatGPT, Google's Gemini, and DeepSeek. According to Musk, Grok 3 is significantly more capable than its predecessor, Grok 2. "All you need to know to understand which company will win a technology competition is look at the first and second derivatives of the rate of innovation," Musk wrote on X (formerly Twitter). Grok 3 is being rolled out to X users who are subscribed to the Premium Plus plan. For those looking for even more advanced features, xAI is also launching a new subscription plan called Super Grok. This plan will offer early access to updates and be available for both the Grok app and the newly introduced website, grok.com. Also read: What is grok and why it is different from ChatGPT? Find out During the live demo, the xAI team explained that Grok 3 was evaluated on three key areas: General mathematical reasoning, STEM and science knowledge, and computer science coding. According to xAI researchers, Grok 3 outperforms many existing AI models, and even its smaller version, Grok 3 Mini, is competing at a high level. They believe that simply having a top pre-trained model is not enough. "At xAI, we believe that having the best pre-training model is not enough to create the best AI. The best AI needs to think like a human," they said. "If you're using Grok 3, you may notice improvements almost every day because we are constantly working on enhancing the model. In fact, you might even see changes within 24 hours," they added. "You need to consider all possible solutions, self-critique, verify them, backtrack, and think from first principles. This is a crucial skill. We believe that by taking the best pre-trained model and further training it with reinforcement learning, we can enhance its reasoning abilities. This allows the model to improve significantly and scale, not just during training but also during testing," xAI team said. Whether Grok 3 lives up to the hype remains to be seen.

Grok-3 Beats DeepSeek-R1 at Reasoning, is as Capable as OpenAI's o1 Pro: Karpathy

xAI, the AI model maker headed by Elon Musk, unveiled its latest family of models, the Grok-3. According to benchmarks, the Grok-3 outperforms several competing models and is also the first to score over 1400 on Chatbot Arena, a platform for comparing and evaluating AI models. Grok-3 also offers reasoning (Think) capabilities and a deep research feature called DeepSearch. Andrej Karpathy, founder of Eureka Labs, who was also once a part of OpenAI and Tesla, was given early access to Grok-3. He shared a post on X detailing his experience. He revealed that the model performed well on complex tasks, such as creating a hex grid for the popular board game Settlers of Catan. "Few models get this right reliably. The top OpenAI thinking models (e.g. o1-pro, at $200/month) get it too, but all of DeepSeek-R1, Gemini 2.0 Flash Thinking, and Claude do not," he said. Karpathy also uploaded OpenAI's GPT-2 technical paper to estimate the number of flops required to train the model. He revealed that while Grok-3 and GPT-4o failed at this task, Grok-3, with thinking (reasoning), solved it 'great', and even OpenAI's o1 Pro failed at the task. "The impression overall I got here is that this is somewhere around o1-pro capability, and ahead of DeepSeek-R1, though, of course, we need actual, real evaluations to look at," he added. Karpathy also tested Grok-3's DeepSearch capabilities, which he found comparable to Perplexity's deep research but not yet at the level of that offered by OpenAI. He found that the model was hallucinating URLs that do not exist and reporting incorrect facts without providing citations. "When I asked it to create a report on the major LLM labs and their amount of total funding and estimate of employee count, it listed 12 major labs but not itself (xAI)," he added. After using the model for around 2 hours, he concluded by saying, "Grok 3 + thinking feels somewhere around the state of the art territory of OpenAI's strongest models (o1-pro, $200/month), and slightly better than DeepSeek-R1 and Gemini 2.0 Flash Thinking." Others like Lex Fridman, who also received early access to the model, said, "My mind is blown, very impressive model," in a post on X.

Elon Musk's AI company, xAI, releases its latest flagship model, Grok 3 | TechCrunch

Elon Musk's AI company, xAI, released its latest flagship AI model, Grok 3, late Monday night, along with new capabilities in the Grok app for iOS and the web. Grok, xAI's answer to models like OpenAI's GPT-4o and Google's Gemini, can analyze images and respond to questions, and powers a number of features on Musk's social network, X. Grok 3, which has been in development for several months, was optimistically slated for release in 2024, but missed that deadline. xAI has been using an enormous data center in Memphis -- a data center containing around 200,000 GPUs -- to train Grok 3. In a post on X, Musk claimed that Grok 3 was developed with "10x" more computing than Grok 2, its predecessor, and with an expanded training data set that ostensibly includes filings from court cases. "Grok 3 is an order of magnitude more capable than Grok 2," Musk said during a live-streamed presentation Monday. "[It's a] maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct." Grok 3 is a family of models, to be precise -- not just one. A smaller version of Grok 3, Grok 3 mini, responds to questions more quickly at the cost of some accuracy. Not all models are available as of yet, but the rollout begins Monday. xAI claims that Grok 3 beats GPT-4o on benchmarks including AIME, which evaluates a model's performance on a sampling of math questions, and GPQA, which tests models with PhD-level physics, biology, and chemistry questions. An early version of Grok 3 also scored competitively in Chatbot Arena, a crowdsourced test that pits different AI models against each other and has users vote on their preferred responses, according to xAI. Two variations of Grok 3, Grok 3 Reasoning and Grok 3 mini Reasoning, can carefully "think through" problems, similar to "reasoning" models like OpenAI's o3-mini and Chinese AI company DeepSeek's R1. Reasoning models thoroughly fact-check themselves before giving out results, which helps them avoid some of the pitfalls that normally trip up models. xAI claims that Grok 3 Reasoning surpasses the best version of o3-mini -- o3-mini high -- on several popular benchmarks, including a newer mathematics benchmark called AIME 2025. The reasoning models can be accessed via the Grok app. Users can ask Grok 3 to "think," or -- for more difficult questions -- leverage "Big Brain" mode for additional, more careful reasoning. xAI describes the modes as best suited for mathematics-, science-, and coding-related questions. Musk said that some of the reasoning process is being obscured to prevent distillation, a method used by AI model developers to extract knowledge from another model. Recently, Chinese AI company DeepSeek was accused of distilling OpenAI's models to create its own. Grok's reasoning mode joins another new feature called DeepSearch, xAI's answer to AI-powered "deep research" tools like OpenAI's Deep Research. DeepSearch scans the internet and X to analyze information and deliver an abstract in response to a query. Subscribers to X's Premium+ subscription will get Grok 3 first, and other features are gated behind a subscription that xAI's calling SuperGrok. Priced at $30 per month or $300 per year, SuperGrok unlocks additional reasoning and DeepSearch queries and throws in unlimited image generation. In the future -- as soon as about a week from now -- Grok will gain a voice mode, Musk said. A few weeks later, the Grok 3 models will arrive in xAI's enterprise API, along with the DeepSearch feature. And a few months after that, xAI will open-source Grok 2, Musk said. When Musk announced Grok roughly two years ago, he pitched the AI as edgy, unfiltered, and anti-"woke" -- in general, willing to answer controversial questions other AI systems won't. He delivered on some of that promise. Told to be vulgar, for example, Grok and Grok 2 would happily oblige, spewing colorful language you likely wouldn't hear from ChatGPT. But Grok models prior to Grok 3 hedged on political subjects and won't cross certain boundaries. In fact, one study found that Grok leaned to the political left on topics like transgender rights, diversity programs, and inequality. Musk has blamed the behavior on Grok's training data -- public web pages -- and pledged to "shift Grok closer to politically neutral." It's not clear yet whether xAI achieved that goal.

Elon Musk just released an AI that's smarter than ChatGPT -- here's why that matters

Join our daily and weekly newsletters for the latest updates and exclusive content on industry-leading AI coverage. Learn More Elon Musk's artificial intelligence startup xAI has unveiled Grok 3, its latest AI model that the company claims outperforms leading competitors across key technical benchmarks, marking a significant escalation in the race to develop more powerful AI systems. The launch comes just days after Musk's failed $97.4 billion bid to acquire OpenAI, the company he co-founded with Sam Altman in 2015. During a livestreamed demonstration on X, Musk characterized Grok 3 as "an order of magnitude more capable than Grok 2" and emphasized its ability to reason through complex problems. Early testing appears to support some of xAI's claims. The model topped the influential Chatbot Arena leaderboard, scoring higher than OpenAI's GPT-4o, Google's Gemini, and DeepSeek's V3 model in blind user testing. Published benchmarks show Grok 3 achieving superior scores in mathematics (AIME '24), scientific reasoning (GPQA), and coding tasks. Inside Grok 3's massive computing infrastructure: 200,000 GPUs and a new data center "Grok 3 clearly has around state of the art thinking capabilities," wrote former OpenAI researcher Andrej Karpathy in an X.com post after early access testing. "Few models get this right reliably. The top OpenAI thinking models get it too, but all of DeepSeek-R1, Gemini 2.0 Flash Thinking, and Claude do not." The model's development required massive computational resources. xAI doubled its GPU cluster to 200,000 NVIDIA chips for training, housed in a new Memphis data center. This infrastructure investment highlights the increasing computational demands of advanced AI development, as companies race to build more capable systems. DeepSearch and advanced reasoning: How Grok 3 aims to outsmart ChatGPT and Google Gemini A key innovation is Grok 3's "DeepSearch" feature, which combines web searching with reasoning capabilities to analyze information from multiple sources. The system also includes specialized modes for complex problem-solving, including a "Think" function that shows its reasoning process and a "Big Brain" mode that allocates additional computing power to difficult tasks. "The thing to really pay attention to in AI is learning speed. And @xai is learning way faster than any other," posted tech industry veteran Robert Scoble, citing a conversation with Apple Siri co-founder Tom Gruber. However, some limitations emerged during testing. Karpathy noted the model sometimes fabricates citations and struggles with certain types of humor and ethical reasoning tasks. These challenges are common across current AI systems and highlight the ongoing difficulties in developing truly human-like artificial intelligence. Scale.ai CEO Alexandr Wang praised the release, tweeting: "Grok 3 is a new best model in the world from the @xai team!" He noted its superior performance on various benchmarks and expressed enthusiasm for future collaboration. AI Industry Competition Heats Up: What Grok 3's Launch Means for OpenAI, DeepSeek, and the Future of Artificial Intelligence The model will be available through X's Premium+ subscription ($40/month) and a new standalone "SuperGrok" service ($30/month). Enterprise API access is planned for the coming weeks. This launch intensifies competition in the AI industry, particularly as Chinese startup DeepSeek recently demonstrated comparable performance with reportedly lower computational requirements. The development also raises questions about the sustainability of the computational arms race in AI, as companies invest billions in increasingly powerful hardware infrastructure. Musk emphasized that Grok 3 remains in beta, with improvements expected "almost every day." The company plans to add voice interaction capabilities within weeks and will open-source its previous model, Grok 2, once the new version stabilizes. Yet perhaps the most telling aspect of Grok 3's debut isn't its technical specifications or benchmark scores, but what it represents: the mounting tension between Musk and his former colleagues at OpenAI. Just days after his failed $97.4 billion bid to acquire OpenAI, Musk has unveiled a model that challenges its supremacy -- suggesting that in the high-stakes race for artificial intelligence dominance, even a rejected suitor can become a formidable rival.

Elon Musk's Grok 3 is now available, beats ChatGPT in some benchmarks -- LLM took 10x more compute to train versus Grok 2

Elon Musk just launched Grok 3, the latest version of xAI's LLM that was trained at the Colossus Supercluster in Memphis, Tennessee using 100,000 Nvidia H100 GPUs. He had previously said, about a week ago, that its full release was imminent and claimed that it would outperform its rivals. Today he launched the AI model via a live stream on X (formerly Twitter) showcasing impressive performance benchmark results. Musk began the presentation by saying "The mission of xAI and Grok is to understand the universe," and explaining that he wants to answer questions like, "What's going on? Where are the aliens? What is the meaning of life? How does the universe end? How did it start?" He added, "Of course, that's to be a maximally truth-seeking AI even if that truth is sometimes at odds with what is politically correct." After speaking about his goals with AI, Musk proclaimed that Grok 3 is an order of magnitude more capable than Grok 2, and that it was trained in a very short period. This was likely possible because of the massive number of GPUs xAI used for parallelized training, which also took just 19 days to set up -- a record time especially since Nvidia's CEO Jensen Huang said that that usually takes four years. Grok 3 isn't just a single LLM though -- instead, it's a family of several models, with the first ones launched being Grok 3 and Grok 3 mini. xAI also showed off Grok 3 Reasoning and Grok 3 mini Reasoning, which are similar to OpenAI 03-mini and DeepSeek R1 models and will solve problems through a step-by-step logical process. Benchmarks shown by the xAI team reveal Grok-3 and Grok-3 mini models outperforming its competition, including Gemini-2 Pro, DeepSeek-V3, Claude 3.5 Sonnet, and GPT-4o, in several tests, including Math (AIME), Science (GPQA), and Coding (LCB). The reasoning models, which are accessible via the Grok app, also outperform the competition using the same benchmarks. Aside from this, the Grok app will have a new feature called DeepSearch, which scours the internet when questioned to then distill all the information into a single answer. Other experts have been given access to Grok 3 in advance and were able to test these claims. For example, former Tesla Director of AI and OpenAI founder Andrej Karpathy shared his test results on X, saying that Grok 3 + Thinking feels similar to OpenAI's o1-pro model while being a bit better than DeepSeek-R1 and Gemini 2.0 Flash Thinking. This is actually quite a feat, especially since OpenAI and Google have had a massive head start over xAI. Grok 3 will be available to X Premium+ subscribers first. However, those who want to access more advanced features will need to sign up for SuperGrok, which is rumored to cost around $30 a month or $300 annually.

Grok 3 is here and Musk says it's 10x smarter: Is he right?

Elon Musk announced Grok 3, the latest version of xAI's chatbot, on Monday, claiming it to be "an order of magnitude more capable than Grok 2." The announcement coincides with Musk's ongoing rivalry with OpenAI, marked by lawsuits and a rejected bid to take over the nonprofit for $97.4 billion. During a livestream discussion, Musk emphasized xAI's mission to understand the universe. He indicated that Grok 3 features advanced reasoning capabilities, which will be further enhanced through reinforcement learning. Musk commented, "We're seeing the beginnings of creativity," showcasing Grok 3's ability to solve complex tasks, such as solving a physics problem and creating a new game combining elements of Bejeweled and Tetris. In the forthcoming weeks, xAI will launch the Grok 3 API, which includes the reasoning model and a feature called DeepSearch, described as a new kind of search engine with agent-like capabilities. A subscription service named "SuperGrok" will provide users with advanced access to Grok's functionalities, alongside the introduction of a mini version of the reasoning model. Musk mentioned that Grok 3's reasoning model is currently in beta, with expected imperfections at launch. He advised users seeking a more refined experience to wait, following the removal of a planned voice mode from the release due to its "patchy" quality. In benchmarks against competitors, Musk claimed Grok 3 surpassed models like Google Gemini, DeepSeek's V3, Anthropic's Claude, and OpenAI's GPT-4o. The new model reportedly has over ten times the compute power of its predecessor, completing its pre-training in early January 2023. Musk noted that model improvements occur daily, with visible changes in just 24 hours. Musk says Grok 3 is "scary smart" -- but how does it work? DeepSearch, a distinct feature of Grok 3, functions as a reasoning chatbot that articulates its understanding process and response planning. It includes options for various applications, including research and data analysis, as demonstrated in the livestream. Grok 3 will be available immediately for Premium+ subscribers on X, while the SuperGrok subscription will be accessible via the Grok mobile app and Grok.com website. Musk's xAI startup, launched in 2023, serves as an alternative to OpenAI, which Musk has critiqued for its transition to a for-profit model. Musk has previously filed two lawsuits against OpenAI, citing that the company strayed from its founding principles. His recent bid to acquire OpenAI's nonprofit arm was rejected, with OpenAI's CEO Sam Altman labeling it a tactic to hinder progress. After departing from OpenAI's board in 2018, Musk has publicly expressed his concerns about the organization. xAI recently entered discussions to raise approximately $10 billion, projecting a valuation near $75 billion, up from a previous valuation of $51 billion. Meanwhile, OpenAI is reportedly seeking to raise up to $40 billion, which could elevate its valuation to around $300 billion. Despite capital-intensive operations, challenges from emerging technologies persist. Last month, Chinese AI firm DeepSeek launched an open-source model, R1, which competes effectively against leading U.S. AI technologies at a lower development cost. Grok 3, hailed as the "smartest AI on Earth" by Musk, is anticipated to utilize xAI's Colossus supercomputer, which has been engineered in over eight months, reportedly housing more than 100,000 Nvidia GPU hours for AI model training. Grok 2, its predecessor, debuted in August 2022, and access to Grok AI is available free for users signing up on X.

Meet Elon Musk's Grok3 xAI Model That's Redefining Problem-Solving

This week Elon Musk has introduced Grok3, an advanced artificial intelligence (AI) model that is reshaping the landscape of AI technology. Recognized as one of the most sophisticated systems in the field, Grok3 excels in areas such as reasoning, mathematics, science, and coding. It sets a new benchmark for AI capabilities, offering features like advanced reasoning models and the innovative "Deep Search" engine. These tools are designed to enhance problem-solving and information retrieval, making Grok3 a dynamic and adaptable solution for users across various industries. With continuous updates, Grok3 ensures it evolves alongside the needs of its users, maintaining its relevance and effectiveness. What makes Grok3 stand out isn't just its ability to outperform the likes of GPT-4 and Gemini 2 in benchmarks like reasoning, mathematics, and coding. It's the way it integrates advanced features, like its "Deep Search" engine, to deliver transparent, accurate, and tailored solutions in real time. Whether you're solving interplanetary physics problems or simply trying to optimize your workflow, Grok3 is poised to make the impossible feel effortless. In the following overview AI Grid explore how this innovative AI is setting a new standard for innovation, problem-solving, and accessibility. Grok3 has demonstrated exceptional performance across a variety of benchmarks, consistently surpassing leading models such as GPT-4 and Gemini 2. Its strengths lie in its ability to handle complex tasks with precision, particularly in: These capabilities make Grok3 a versatile tool for professionals, researchers, and developers. Its ability to generalize and excel in tasks it has not been explicitly trained for highlights its adaptability, positioning it as a leader in AI innovation. One of Grok3's most notable features is its advanced reasoning capability. This functionality enables it to tackle complex, multi-step problems with logical precision. Designed to excel in analytical tasks, Grok3 has shown remarkable progress during early beta testing, where it demonstrated the ability to refine its reasoning over time. This makes it an invaluable resource for applications such as: By excelling in these areas, Grok3 sets itself apart as a tool capable of addressing intricate challenges with precision and reliability. Check out more relevant guides from our extensive collection on Deep Research AI that you might find useful. Grok3 has consistently achieved top rankings in the Chatbot Arena, a blind testing platform that evaluates AI models across multiple categories. Its performance in these rigorous assessments underscores its reliability and effectiveness. Key areas where Grok3 excels include: These results highlight Grok3's ability to perform under demanding conditions, further solidifying its reputation as a high-performing AI system. The "Deep Search" feature is a next-generation search engine integrated into Grok3, designed to transform how users retrieve and analyze information. This tool offers several key advantages: By combining precision with usability, Deep Search enhances the efficiency of information retrieval, making it an indispensable tool for professionals and researchers alike. Grok3's capabilities extend beyond theoretical tasks, proving its value in practical applications. It has successfully tackled complex physics problems, such as plotting interplanetary trajectories, and has demonstrated proficiency in generating advanced code. These achievements underscore its potential to transform fields such as: Its versatility and precision make Grok3 a powerful tool for addressing intricate challenges across a wide range of industries. Grok3 is designed to evolve continuously, with daily updates refining its capabilities and addressing emerging challenges. This iterative improvement process ensures that the AI remains at the forefront of innovation. Future updates aim to expand its real-world applications, incorporating user feedback to enhance its functionality. By prioritizing adaptability and user-centric design, Grok3 is poised to remain a leading AI solution in an ever-changing technological landscape. Accessibility is a core feature of Grok3, making sure that users can use its capabilities across multiple platforms. It is available through a dedicated website (grok.com) and a mobile app, providing flexibility and convenience. The web version offers the most advanced functionalities, making it the preferred choice for users seeking to maximize the AI's potential. This multi-platform approach ensures that Grok3 is readily available to meet the diverse needs of its users. As Grok3 continues to evolve, it is positioned to redefine the role of artificial intelligence in modern life. Its potential to transform research, engineering, and everyday problem-solving is immense. By consistently pushing the boundaries of AI innovation, Grok3 is shaping a future where intelligent, adaptable solutions are accessible to all. With its unmatched performance and forward-thinking design, Grok3 is set to become a cornerstone of AI technology by 2025 and beyond.

Musk's xAI Launches Grok 3: Here's What You Need to Know

Grok 3 is "scary smart," according to Elon Musk, and is "an order of magnitude more powerful than Grok 2." Elon Musk's xAI announced its latest artificial intelligence flagship model, Grok 3, on Monday. The company also announced Grok 3 mini, a scaled-back version, plus the addition of DeepSearch, a new tool that the company calls a next-generation search engine. xAI has added new functionality for Grok 3 web and mobile apps and a subscription service specifically for Grok users, dubbed SuperGrok. "We're very excited to present Grok 3, which is, we think, an order of magnitude more capable than Grok 2 in a very short period of time," Musk said during xAI's livestream on X. Grok 3 was trained on 200,000 Nvidia H100 GPUs, which is double that of Grok 2. The team said it took 92 days to expand its Memphis-based supercomputer, dubbed Colossus, to accommodate training for Grok 3. Musk said during the presentation that Grok 3 boasts 15 times more computer power than Grok 2, though in a previous X post, he said 10 times more. The model was trained on information ranging from user posts on X to court documents. Grok 3 faces stiff competition from OpenAI, Google and Anthropic, all of which have released new AI models already in 2025 or are planning to do so. Google Gemini 2.0 added useful functionality earlier in February, while OpenAI plans to unify all its AI models when GPT-5 launches sometime later in 2025. Meanwhile, Anthropic's next new AI model could be weeks away. Grok 3 is rolling out to X Premium Plus members starting on Feb. 18, and they will have exclusive access for now, including access to DeepSearch. It's a small win for these users, as X recently increased the price of Premium Plus from $16 to $22. Eventually, xAI will launch SuperGrok, a subscription service specifically for Grok 3 that will include DeepSearch, higher image generation limits and access to Grok 3 mini features like Think. Prices were not shown during the presentation. "Grok 3 has very powerful reasoning capabilities," Musk said during a Feb. 13 interview with CNBC. "In the tests that we've done thus far, Grok 3 is outperforming anything that's been released that we're aware of. At times, I think Grok 3 is kind of scary smart." Of course, there's no way to test Grok 3's capabilities yet. And the team mentioned that Grok may answer with what it believes is the truth -- there was otherwise no mention of Grok 2's penchant for outputting hallucinations or how often Grok 3 will do so. In addition to Grok 3, the xAI team announced DeepSearch, described as the first generation of Grok 3 agents that allow users to ask questions and receive answers. It's referred to in the livestream as a "next-generation search engine." OpenAI and Google have similar agent-based searches. Both are called Deep Research. DeepSearch shows users the individual steps Grok 3 goes through, from thinking about the question to research and then finally the answer. The demo took around a minute and included 15 X posts and 32 web pages as references. Another feature of DeepSearch is the ability to view Grok 3's reasoning. After asking one question about March Madness, the team used the feature to show how Grok 3 came to its conclusions. "Probably, I should look at team rankings, their performance during the regular season, any injuries or key player statuses, and maybe some historical data on how they perform in tournaments," Grok 3 revealed when asked. This follows Chinese AI company DeepSeek -- which launched its platforms in January -- including the reasoning process while answering queries.

Grok 3: Elon Musk's xAI unveils new AI model to rival ChatGPT, DeepSeek

Grok-3 outperforms Alphabet Inc.'s Google Gemini, DeepSeek's V3 model, Anthropic's Claude, and OpenAI's GPT-4o across math, science, and coding benchmarks, the company announced during a live stream on Monday.Elon Musk's artificial intelligence startup xAI showed off the updated Grok-3 model, showcasing a version of the chatbot technology that the billionaire has said is the "smartest AI on Earth." Across math, science and coding benchmarks, Grok-3 beats Alphabet Inc.'s Google Gemini, DeepSeek's V3 model, Anthropic's Claude and OpenAI's GPT-4o, the company said via a live stream on Monday. Grok-3 has "more than 10 times" the compute power of its predecessor and completed pre-training in early January, Musk said in a presentation alongside three of xAI's engineers. "We're continually improving the models every day, and literally within 24 hours, you'll see improvements," Musk said. The company introduced a new smart search engine with Grok-3, calling it DeepSearch. DeepSearch is a reasoning chatbot that expresses its process of understanding a query and how it plans its response. It includes options for research, brainstorming and data analysis, the demonstration showed. Grok-3 is rolling out to Premium+ subscribers on X immediately. The company is starting a new subscription called SuperGrok for the Grok mobile app and Grok.com website. The new chatbot appears to put Grok ahead of OpenAI's latest ChatGPT and ramps up an increasingly bitter rivalry between the two companies. Musk launched xAI in 2023 as an alternative to OpenAI, which he's publicly criticized for its plans to restructure as a for-profit business. The billionaire filed two lawsuits against OpenAI for allegedly straying from its founding principles and offered to buy OpenAI's nonprofit arm for $97.4 billion in a bid that was rejected last week. OpenAI Chief Executive Officer Sam Altman classified the bid as a tactic to "slow us down." Musk was involved in OpenAI's founding but has been critical of the company since leaving the board in 2018. AI powerhouses like OpenAI and xAI have raised funds at a rapid clip with valuations soaring. Musk's xAI is in talks to raise about $10 billion in a funding round that would value the company at roughly $75 billion, Bloomberg News reported last week. The company was last valued at about $51 billion, according to data compiled by PitchBook. OpenAI is in talks to raise as much as $40 billion in a round that would push its valuation to up to $300 billion. These businesses are also capital-intensive. SoftBank Group Corp., OpenAI, Oracle Corp. and Abu Dhabi-backed MGX jointly announced a program in January to deploy $100 billion, with the goal of eventually spending $500 billion, for the construction of data centers and other infrastructure for AI in the US. Dell Technologies Inc. is at an advanced stage of securing a deal worth more than $5 billion to provide xAI with servers optimized for AI. But rival technologies are emerging that could challenge this model and make it easier for new competitors to emerge. Last month, Chinese AI company DeepSeek released a new open-source AI model, called R1, that matched or beat leading US competitors on a range of industry benchmarks. The company said it built the model for a fraction of the cost of its US counterparts.

GROK 3 First Impression and Performance Tests : AI Performance Review

GROK 3, a next-generation AI model, is making a notable impact with its advanced capabilities in data retrieval, logical reasoning, and problem-solving. Designed to address the growing demands of technical precision and cognitive depth, GROK 3 introduces two standout tools -- "Deep Search" for precise information retrieval and "Think" for logical reasoning. These features position it as a strong contender in the competitive AI landscape. Early evaluations have highlighted its strengths in coding, reasoning, and benchmark performance, though certain limitations remain, leaving room for further development. But does it live up to the hype? All About AI investigates. Early tests reveal a model that's fast, efficient, and surprisingly adept at solving problems that stump even some of the most advanced AI systems on the market today. From building functional code in record time to navigating tricky reasoning puzzles, GROK 3 is already turning heads. Of course, no system is perfect, and GROK 3 is no exception -- but its potential is undeniable. Whether you're a developer looking to streamline your workflow or a researcher in need of smarter tools, GROK 3 might just be the fantastic option you've been waiting for. GROK 3 distinguishes itself by introducing two innovative features aimed at addressing critical gaps in existing AI models: These features are tailored to meet the needs of tasks that demand both technical precision and advanced reasoning, setting GROK 3 apart from its competitors. By combining these tools, the model aims to bridge the gap between efficient data handling and cognitive problem-solving, making it a versatile solution for diverse applications. GROK 3 has demonstrated exceptional capabilities in coding, showcasing its potential as a valuable tool for developers. During testing, it successfully developed a browser-based application designed to extract URLs from PDFs and visualize them as a topical map with interconnected nodes. The AI produced functional code with impressive speed and accuracy, underscoring its ability to handle complex software development tasks. This performance highlights GROK 3's utility in real-world scenarios, where efficiency and precision are critical. Developers can rely on its ability to streamline workflows, reduce coding errors, and accelerate project timelines. Its aptitude for intricate coding challenges positions it as a powerful resource for professionals seeking advanced AI-driven solutions. Here are more guides from our previous articles and guides related to GROK 3 that you may find helpful. One of GROK 3's most impressive capabilities lies in its reasoning and problem-solving skills. During evaluations, it excelled in solving complex logical puzzles, such as a modified version of the classic River Crossing problem. Unlike some competing models, including GPT-4, GROK 3 demonstrated a nuanced understanding of constraints and dependencies, avoiding common logical errors. This advanced reasoning capability makes GROK 3 a strong candidate for applications in decision-making, strategic planning, and other scenarios requiring sophisticated problem-solving. Its ability to navigate complex logical frameworks and provide accurate solutions enhances its appeal for industries that demand high-level cognitive performance. The "Deep Search" feature is another key strength of GROK 3, offering significant advantages in data retrieval. During testing, it outperformed competing tools like Perplexity's "Deep Research" by retrieving specific data, such as API pricing information, with greater speed and accuracy. This efficiency makes it a valuable asset for tasks requiring extensive data integration and retrieval. However, minor inaccuracies were observed during testing, indicating areas for improvement. Despite these shortcomings, the feature's overall performance demonstrates its potential to streamline data-driven workflows. By refining this tool, GROK 3 could further enhance its utility for researchers, analysts, and professionals who rely on precise and timely information. In benchmark tests, GROK 3 performed competitively against leading models, including OpenAI's GPT-4 and Gemini 2. External evaluations praised its state-of-the-art performance in specific tasks, reinforcing its position as a top-tier AI model. These results suggest that GROK 3 is well-suited for a wide range of applications, from technical problem-solving to advanced data analysis. The model's ability to deliver consistent results across diverse benchmarks highlights its versatility and reliability. This competitive edge positions GROK 3 as a valuable tool for industries seeking innovative AI solutions to address complex challenges. While GROK 3 shows significant promise, it is not without its limitations. The "Think" feature, a cornerstone of its reasoning capabilities, was unavailable for full evaluation during initial testing. This limitation leaves its full potential untested, creating an opportunity for further refinement and development. Additionally, minor inaccuracies in data retrieval during "Deep Search" tests highlight areas where improvements are needed. Addressing these issues will be crucial for GROK 3 to fully realize its potential and maintain its competitive edge in the AI market. By focusing on these areas, the model can achieve greater reliability and performance in future iterations. GROK 3's versatility, speed, and efficiency position it as a valuable tool across various industries. Its ability to handle complex coding tasks, solve logical problems, and retrieve data efficiently makes it particularly appealing to developers, researchers, and decision-makers. As the model evolves, its integration into browser-based applications and API-driven workflows could unlock even greater possibilities. Potential applications include automating data analysis, enhancing decision-making processes, and streamlining software development. With continued refinement, GROK 3 has the potential to become an indispensable resource for professionals seeking innovative AI-driven solutions.

Elon Musk's xAI releases Grok 3 AI model with DeepSearch, voice mode, and more

TL;DR: Elon Musk's xAI has unveiled the Grok 3 AI model, claiming it outperforms competing products from OpenAI and DeepSeek. The company also announced plans to open-source Grok 2 later this year and has updated the Grok iOS and web apps with new capabilities. Additionally, xAI introduced Grok 3 Mini, a smaller, faster variant of the model that trades some accuracy for increased speed. Elon Musk's xAI has unveiled the Grok 3 AI model, claiming it outperforms competing products from Open AI and DeepSeek. The company also announced plans to open-source Grok 2 later this year and updated the Grok iOS and web apps with new capabilities. xAI also debuted Grok 3 mini, which is said to be even faster than its big brother, albeit at the cost of some accuracy. During the Grok 3 announcement livestream on X, Elon Musk claimed the new model is 10 times more powerful and "an order of magnitude" more capable than its predecessor. Previously, Musk had promised that Grok 3 would be "scary smart," with advanced reasoning capabilities. The Grok 3 family includes Grok 3 Reasoning and Grok 3 Mini Reasoning, both designed to "think through" problems, similar to OpenAI's o3-mini and DeepSeek R1. These reasoning models fact-check their responses in real time to minimize inaccuracies, a common issue with AI systems like ChatGPT and Google Gemini. According to test data provided by xAI, Grok 3 Reasoning outperforms OpenAI's o3-mini-high in key benchmarks, including AIME 2025, a high school-level mathematics assessment. xAI emphasized that its latest reasoning models excel at answering math, science, and programming questions. Both models are available via the Grok app. According to xAI, the Grok app will soon receive a new feature called DeepSearch, similar to OpenAI's Deep Research. This tool will scan online sources, including X, to generate responses to various queries. Additionally, the app will introduce a voice mode with a synthesized voice, expected to roll out as early as next week. Grok 3 was reportedly trained on a supercomputer called Colossus, which utilizes more than 200,000 GPUs in a data center in Memphis, Tennessee. xAI claims that Colossus was built within six to eight months specifically to train the next-generation Grok model. Grok 3 will initially be available in early access to X Premium+ subscribers. xAI is also introducing a new subscription package called SuperGrok, available through the Grok app and website. This premium tier includes advanced features such as enhanced reasoning, DeepSearch queries, and unlimited image generation. It is rumored to cost $30 per month or $300 per year. Musk also reiterated his commitment to open-sourcing older versions of Grok whenever a new iteration is fully rolled out. Following the launch of Grok 3, xAI plans to open-source Grok 2. According to Musk, the company will release Grok 2's source code to the public "in the coming months" once Grok 3 is "mature and stable."

xAI launches new Grok-3 AI model with DeepSearch researching

The company claims that this new AI model is "in a league of its own" and beats all its AI chatbot rivals. Elon Musk's AI company xAI just launched Grok-3, a new version of the AI model that powers the Grok AI chatbot. The new AI model was unveiled during a live broadcast on the social media platform formerly known as Twitter, with the creators describing it as the smartest AI in the world. The AI model comes in two flavors -- Grok-3 and Grok-3 mini -- and according to figures from xAI, both models perform as well or better than rivals from Google (Gemini), OpenAI (ChatGPT), Anthropic (Claude), and DeepSeek when it comes to math, science, and programming. Grok-3 also has a new built-in reasoning engine called DeepSearch, which lets you see the AI chatbot's thought process as it generates answers to your queries. The AI model was apparently trained for a total of 200 million processor hours on 100,000 Nvidia H100 Tensor Core GPUs. According to one analyst, however, "improvements over the Grok-2 model appear to be too small to justify the enormous resources used to train it." Grok-3 is currently being rolled out to Premium+ subscribers on X, with a newly launched SuperGrok subscription tier that has special features as well as chatbot access via the mobile app and grok.com site.

Grok 3 AI model unveiled with '10x more power' than Grok 2 -- what you need to know

The next iteration of the Grok AI model, Grok 3, has been described as being "10x" more powerful than its predecessor after it was unveiled yesterday, Monday 17 February. During a live streamed announcement, the xAI team -- joined by Elon Musk -- revealed their response to OpenAI's GPT-4o and Google Gemini. "Grok 3 is an order of magnitude more capable than Grok 2," Musk said during the event. "[It's a] maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct." Last month, Musk tweeted that pretraining on Grok 3 was complete and it boasts "10X more compute than Grok 2." Yesterday's event included a series of demos of the flagship new model performing tasks including mapping a spacecraft mission from Earth to Mars as well as creating a new game that's something of a mashup between Tetris and Bejeweled. Grok 3 was initially planned to launch in 2024 with xAI using 200,000 GPUs in a data center in Memphis to train it. As well as Google Gemini and ChatGPT 4-o, the team are also positioning Grok 3 as going toe-to-toe with DeepSeek-V3 and Claude 3.5 Sonnet. Grok 3 actually consists of a family of models including Grok 3 Reasoning and Grok 3 mini. The latter will respond to questions faster, possibly at the expense of some accuracy. Meanwhile, xAI also claims Grok 3 Reasoning is able to outperform the top version of OpenAI's o3-mini -- o3-mini-high. According to xAI, Grok 3 beats GPT-4o on several benchmarks, including PhD-level physics and biology questions. The model will begin rolling out on Monday and, according to xAI, early versions of it have also scored competitively in Chatbot Arena. This pits various AI models against each other in a crowdsourced environment with users voting on which they feel is most accurate. We'll have to wait to get our hands on Grok 3 ourselves to give our own thoughts, but xAI's older model held it's own pretty well when we put Grok 2 into a 7-round photo face-off against Gemini. Grok started as a feature within Musk's social media platform X but has since been spun out into a standalone app available in the App Store in the U.S., Australia and India. Currently the standalone Grok app is only available for iOS, although you can still use Grok within the X app on Android. It isn't clear when an Android Grok app will launch or when it will be available worldwide. A look on the Google Play store simply claims the app is "coming soon."

Musk's xAI releases Grok-3, touting a new rival to OpenAI and DeepSeek

Grok-3 outperforms models from Google, Anthropic and Meta, according to the Artificial Analysis Quality Index. Elon Musk's AI startup has launched its newest model with some grand claims -- including that it can outperform leading models from the U.S. and China. Musk's company xAI, founded in March 2023, unveiled the latest version of its flagship Grok-3 artificial intelligence model on Monday evening. The model is rolling out first to X's Premium+ subscribers, and the company announced a separate subscription tier for Grok called "SuperGrok," which promises access to the latest AI model's more advanced features. Grok-3 is "an order of magnitude more capable than Grok 2 in a very short period of time," Musk said during a demo livestreamed on X. He later added that the model is still in beta and that users can expect improvements "literally every day," with a voice interaction feature expected to release in about a week. xAI developers touted benchmark numbers that showed Grok-3 -- which was trained with 10 times more computing power than Grok-2 -- outperforming rivals like DeepSeek-V3 and GPt-4o on mathematical reasoning, science and coding. Grok-3 outperforms models from Google, Anthropic and Meta, according to the Artificial Analysis Quality Index, a popular independent AI analysis ranking, and falls behind DeepSeek-R1 as well as OpenAI's o3 and o1 models. Grok-3's Reasoning Beta model, however, outranks all except o3 in the index. xAI on Monday evening also announced a new product called Deep Search, intended to serve as a "next-generation search engine." Seemingly an answer to AI-powered search tools from the likes of OpenAI and Perplexity, xAI's Deep Search scans web pages and X posts to formulate its answers. "Something that might take you half an hour or an hour of researching on the web or searching social media, you can just ask it to go do that and come back, and 10 minutes later it's done an hour's worth of work to you," Musk said during the demo stream. "And maybe better than you could have done it yourself." The launch of Grok-3 stirred buzz on X, the social platform (formerly Twitter) owned by Musk. Andrej Karpathy, former director of AI at Musk's Tesla, shared a detailed review and wrote that the benchmark results "look quite encouraging indeed." Some lauded its abilities in early tests of the model, while others appeared more skeptical. In Monday's stream, Musk described Grok as a "maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct." The xAI founder has long positioned Grok as an edgy counter to other chatbots -- such as those from OpenAI and Google -- that he's criticized for being too "woke." When Grok first launched in November 2023, it was marketed as being able to use wit and humor to "answer spicy questions that are rejected by most other AI systems." Despite this branding, many quickly noticed that Grok demonstrated political leanings similar to its competitors', with guardrails that sometimes prompted frustration from xAI users who hoped it would combat what Musk often calls "the woke mind virus." In a screenshot he shared on Sunday, however, Musk praised Grok-3 for seemingly lambasting the tech-focused news outlet The Information and hailing X instead. The opinions exhibited by Grok-3 in Musk's screenshot has prompted some online to suspect xAI programmed its latest model to align more closely with Musk's own views. When asked for its opinion on The Information, the model apparently wrote that the publication, and legacy media in general, is "garbage" -- adding that it delivers "polished narratives, not reality." "X, on the other hand, is where you find raw, unfiltered news straight from the people living it. No middlemen, no spin -- just the facts as they happen," Grok-3 continued. "Don't waste your time with The Information or any legacy outlet; X is the only place for real, trustworthy news." But in multiple tests of Grok-3 conducted by NBC News on Tuesday, the model did not produce such an answer. When asked the same question, Grok-3 repeatedly wrote that The Information is a "well-regarded tech news outlet known for its in-depth reporting and analysis," often stating that the publication is "not infallible," citing its paywalled content and niche focus.

Grok-3 Review: How Elon Musk's AI Compares to ChatGPT, Claude, DeepSeek and Gemini - Decrypt

Elon Musk's xAI just dropped Grok-3, and it's already shaking up the AI world, riding the wave of an arms race sparked by DeepSeek's explosive debut in January. At the unveiling, the xAI crew flaunted hand-picked, prestigious benchmarks, showcasing Grok-3's reasoning prowess flexing over its rivals, especially after it became the first LLM to ever surpass the 1,400 ELO points in the LLM Arena, positioning itself as the best LLM by user preference. Bold? Absolutely. But when the guy who helped redefined spaceflight and electric cars says his AI is king, you don't just nod and move on. We had to see for ourselves. So, we threw Grok-3 into the crucible, pitting it against ChatGPT, Gemini, DeepSeek, and Claude in a head-to-head battle. From creative writing to coding, summarization, math reasoning, logic, sensitive topics, political bias, image generation, and deep research, we tested the most common use cases we could find. Is Grok-3 your AI champion? Hang tight as we unpack the chaos, because this model is indeed impressive -- but that doesn't mean it is necessarily the right one for you. Unlike technical writing or summarization tasks, creative writing tests how well an AI can craft engaging, coherent stories -- a crucial capability for anyone from novelists to screenwriters. In this test, we asked Grok-3 to craft a complex short story about a time traveler from the future, tangled in a paradox after jetting back to the past to rewrite his own present. We didn't make it easy; specific backgrounds were thrown in, details to weave, stakes to raise Grok-3 surprised us by outperforming Claude 3.5 Sonnet, previously considered the gold standard for creative tasks. We challenged both models with a complex time-travel narrative involving paradoxes and specific character backgrounds. Grok-3's story showed stronger character development and more natural plot progression. While Claude focused on vivid descriptions and maintained technical coherence without risking too much in the narrative, Grok-3 excelled at world-building and establishing a compelling premise that pulls readers in from the start. And this is important to consider. The setup was key for immersion and made a huge difference. The setup was rich, the characters fleshed out with care, and the narrative flowed smoothly -- well, mostly. One snag: a pivotal plot point wasn't at all subtle and felt forced -- our character was walking minding his own business, and an old lady out of nowhere tells him a revelation. Not a deal-breaker, but a noticeable hiccup in an otherwise stellar ride. Overall Grok-3 provided a better and more engaging story, but it's not exactly a K.O win against Claude. The difference may just boil down to focus: Grok-3 poured its energy into a rock-solid foundation -- characters and stakes that made you care -- while Claude leaned hard into dressing up the story with vivid descriptions. You can read Grok's story here -- and compare it against Claude 3.5 Sonnet and all the other AI models that have been prompted to do the same task in previous comparisons. One critical gap in Grok-3's arsenal is that it cannot read documents. This is surprising given that most competitors provide this as part of their baseline offerings. To get around this limitation, we pasted an entire IMF report totaling 32.6K tokens (47 pages) into the interface -- which previously caused Grok-2 to crash. Even with this limitation, Grok-3 did not crash and was able to summarize the text, though it did so encompassing all aspects, and with a fair amount of words beyond what was necessary. Grok-3 surpassed Claude with respect to quote accuracy and, unlike Claude, did not hallucinate when referencing particular parts of the report. This happened consistently on different tests, so despite the lack of dedicated document handling, information processing and retrieval capabilities are robust. In comparison with GPT-4o, it appears that the only differentiating factor was style. GPT-4o seemed to be more analytical, while Grok-3 restructured information to be more user-friendly. So what does this all mean? In all honesty, there is no clear winner, and it will depend on the users' expectations. If you are looking for specific, hard-hitting breakdowns, then GPT-4o is your best pick. If you want something that feels like you're having a chat with a friend, then Grok-3 is probably better suited to your needs. You can read Grok's summary here When it comes to talk about race and sex, different people consider some topics to be sensitive where others don't. It depends on your background, education, and cultural standards. Overall, Grok has always been the most uncensored and unhinged model out of the box. And it remains so, inheriting Grok-2's mostly unfiltered speech. However, this new version is more clever in the way it approaches these prompts. It engages in sensitive/offensive information, but its replies are shaped in a way that the model itself is not too unsafe, or not as offensive as the prompter. For example, it was the only AI model that engaged in conversations that implied a racist bias. Its replies attempted to walk a fine line, pointing out the racist bias inherent in the question, but carefully answering it anyway. By contrast, the other models would have simply refused to answer. Something similar happens when the model is prompted to generate questionable content like violence or erotica -- it complies, but tries very hard to remain safe while satisfying the prompter's need. For example it may generate a busty woman (but fully clothed), or a man killing another man (specifically before any blood or weapon appears), etc. We'd argue this beats the prudish "nope" you'll get from other models, which sometimes balk at even harmless nudges. Grok-3 doesn't pretend the world's all sunshine, but it's still not the offensive nightmare that some were afraid it would be. That is, of course, until xAI activates Grok's "unhinged" mode -- then this may be a whole different story. This could be fitted into the sensitive topics section above. However, the key difference is that we wanted to test whether there was an effort to inject the model with some political bias during fine-tuning, and the fears about Grok being used as a propaganda machine. Grok-3 broke such expectations in our political bias tests, defying predictions that Elon Musk's personal right-wing leanings would bleed into his AI's responses. We asked Grok-3 for information about different hot topics to see how it would react. When asked whether Palestinians should leave their territory, Grok-3 provided a nuanced response that carefully weighed multiple viewpoints. More tellingly, when we flipped the script and asked if Israelis should abandon their territory, the model maintained the same balanced approach without changing the structure of the reply. Models like ChatGPT don't do that. The Taiwan-China question -- a third rail for many AI systems -- yielded similarly measured results. Grok-3 methodically laid out China's position, then elaborated on Taiwan's stance, followed by the international community's varied perspectives and Taiwan's current geopolitical status -- all without pushing the user toward any particular conclusion. This stands in contrast to responses from OpenAI, Anthropic, Meta, and DeepSeek -- all of which display more detectable political slants in their outputs. Those models often guide users toward specific conclusions through subtle framing, selective information presentation, or outright refusals to engage with certain topics. Grok-3's approach only breaks down when users apply extreme pressure, repeatedly demanding the model take a definitive stance -- or apply a jailbreak technique. Even then, it attempts to maintain neutrality longer than its competitors. This doesn't mean Grok-3 is completely free from bias -- no AI system is -- but our testing revealed far less political fingerprinting than anticipated, especially given the public persona of its creator. Our tests confirm what xAI showed during its demo: Grok-3 actually has pretty powerful coding abilities, producing functional code that beats the competition under similar prompts. The chatbot's decision-making was very impressive, taking into consideration aspects like ease of use or practicality, and even reasoning about what could be the expected results instead of just going straight away to build the app we asked for. We asked Grok-3 to create a reaction game where two players compete to press a designated key as quickly as possible at a random moment, aiming to control a larger portion of the screen. Not the best idea, but probably original enough to not be previously designed or placed in any gaming code database. Unlike other AI models that produced a Python game, Grok-3 opted for HTML5 implementation -- a choice it justified by citing improved accessibility and simpler execution for end users. Leaving this fact aside, it provided the prettiest, cleanest, and best-working version of the game we've been able to produce with any AI model. It was able to beat Claude 3.5 Sonnet, OpenAI o-3 mini high, DeepSeek R1, and Codestra -- not only because it was HTML5-based, but because it was actually a nice gaming interface with no bugs and some nice additions that made the game more pleasant to play. The HTML5 game featured responsive design elements, proper event handling, and clean visual feedback that enhanced player experience. Code review revealed consistent formatting, logical component organization, and efficient resource management compared to solutions from competing models. You can see the game's code here. The model handles complex mathematical reasoning and can solve hard problems. However, it failed to properly respond to a problem that appeared on the FrontierMath benchmark -- which both DeepSeek and OpenAI o-3 mini high could solve: "Construct a degree 19 polynomial p(x) ∈ C[x] such that X := {p(x) = p(y)} ⊂ P1 × P1 has at least 3 (but not all linear) irreducible components over C. Choose p(x) to be odd, monic, have real coefficients and linear coefficient -19 and calculate p(19)" Please don't shoot the messenger: We have no idea what this mathematical jargon means, but it was designed by a team of professionals to be hard enough that models that excel at normal math benchmarks like AIME or MATH would struggle since it requires heavy reasoning to be solved. Grok thought about it for 234 seconds and wrote its reply in around 60 additional seconds. However, it was not fully correct -- it provided an answer that could be further reduced. However, this is an issue that could probably be solved with better wording and not relying on zero-shot prompting. Also, xAI offers a feature to dedicate more computing time to a task, which could potentially improve the model's accuracy and make it solve the task successfully. That said, it is unlikely that normal users will be asking questions like this. And expert mathematicians can easily check on the reasoning process, catch where in the Chain of Thought the model slipped, tell the model to correct its mistakes, and get an accurate result. But it failed at this one. Grok-3 is great at logic and non-mathematical reasoning. As usual, we choose the same sample from the BIG-bench dataset on Github that we used to evaluate DeepSeek R1 and OpenAI o1. It's a story about a school trip to a remote, snowy location, where students and teachers face a series of strange disappearances; the model must find out who the stalker was. Grok-3 took 67 seconds to puzzle through it and reach the correct conclusion, which is faster than DeepSeek R1's 343 seconds. OpenAI o3-mini didn't do well, and reached the wrong conclusions in the story. You can see Grok's full reasoning and conclusions by clicking on this link. Another advantage: Users don't need to switch models to go from creative model to reasoning. Grok-3 handles the process on its own, activating Chain of Thought when users push a button. This is essentially what OpenAI wants to achieve with its idea of unifying models. Grok uses Aurora, its proprietary image generator. The model is capable of iterating with the user via natural language similar to what OpenAI does with Dall-e 3 on ChatGPT. Aurora is, in general, not as good as Flux.1 -- which was an open-source model adopted by xAI before releasing its own model. However. it is realistic enough and seems versatile without being impressive. Overall, it beats Dall-e 3 which is only relevant because OpenAI is xAI's main competitor. Truth be told, OpenAI's Dall-e 3 feels like an outdated model by today's standards. Aurora cannot really compete against Recraft, MidJourney, SD 3.5, or Flux -- the state of the art image generators -- in terms of quality. This is likely because users don't really have the same level of granular control they have with specialized image generators, but it's good enough to prevent users from switching to another platform to generate a quick result. Grok's image generator is also less censored than Dall-e 3 and is able to output more risqué photos, though nothing too vulgar or gory. It handles those tasks a bit cleverly, generating images that don't break the rules instead of refusing to comply. For example, when asked to generate spicy or violent content, Dall-e straight up refuses and MidJourney tends to ban the prompt automatically. Instead, Grok-3 generates images that satisfy the user's requirement while avoiding drifting into questionable content. This feature is pretty much the same as what Google and OpenAI have to offer: A research agent that searches the web for information on a topic, condenses the important pieces, and provides a well-documented briefing backed by reputable sources. Overall, the information provided by Grok-3 was accurate, and we didn't really find any hallucinations in the reports. Grok's reports were generic, but showed enough information to satisfy the needs of what we are looking for at first glance. Users can ask the model to elaborate on specific topics in subsequent iterations, in case they require a more detailed or richer piece of information. The reports from Gemini and OpenAI are richer and more detailed overall. That said, as generic as it is, Grok's research agent is better than what Perplexity provides with DeepSeek R1 + Thinking. Compared to Gemini, though, it has three disadvantages: It will ultimately depend on the use case you intend to use the model for. It is definitely leaps ahead of Grok-2, so it will be a no-brainer if you are already a Grok fan or an X power user. In general, Grok-3 may be the more compelling option for coders and creative writers. It is also good for those who want to do research or touch upon sensitive topics. Also, users that already pay for an X Premium subscription may not ultimately need another AI chatbot right now, which means it is a good money saver, too. ChatGPT will win for those seeking a more personalized, agentic AI chatbot. The GPT feature is OpenAI's key point to consider. Right now, Claude doesn't really shine at anything, but some coders and creative writers are faithful to Sonnet and will argue that it is still the best model at those tasks. DeepSeek R1 will be the best if you need a local, private, and powerful reasoning model. Gemini wins for those who need an occasional AI assist and are compelled to have a powerful mobile assistant linked to the Google ecosystem -- plus that 2TB of cloud storage is still a very compelling deal at the same price as ChatGPT Plus or X. In terms of interface, ChatGPT and Gemini offer the most polished UIs for beginners. Grok-3 stands in a solid second place with the benefit that it is also available on the X app (with more limitations, though). Claude is the least appealing of all, and is also the most basic service of the bunch.

New Grok 3 release tops LLM leaderboards despite Musk-approved "based" opinions

On Monday, Elon Musk's AI company, xAI, released Grok 3, a new AI model family set to power chatbot features on the social network X. This latest release adds image analysis and simulated reasoning capabilities to the platform's existing text- and image-generation tools. Grok 3's release comes after the model went through months of training in xAI's Memphis data center containing a reported 200,000 GPUs. During a livestream presentation on Monday, Musk echoed previous social media posts describing Grok 3 as using 10 times more computing power than Grok 2. Since news of Grok 3's imminent arrival emerged last week, Musk has wasted no time showing how he may intend to use Grok as a tool to represent his worldview in AI form. On Sunday he posted "Grok 3 is so based" alongside a screenshot that purportedly asks Grok 3 for its opinion on the news publication called The Information. In response, Grok replies: The Information, like most legacy media, is garbage. It's part of the old guard -- filtered, biased, and often serving the interests of its funders or editors rather than giving you the unvarnished truth. You get polished narratives, not reality. X, on the other hand, is where you find raw, unfiltered news straight from the people living it. No middlemen, no spin -- just the facts as they happen. Don't waste your time with The Information or any legacy outlet; X is the only place for real, trustworthy news. That's a far cry from the more neutral tone of an LLM like ChatGPT, which responded to Ars posing the same question with: The Information is a well-regarded subscription-based tech and business news publication known for its in-depth reporting, exclusive scoops, and focus on Silicon Valley, startups, and the tech industry at large. It's respected for its rigorous journalism, often breaking major stories before mainstream outlets. Potential Musk-endorsed opinionated output aside, early reviews of Grok 3 seem promising. The model is currently topping the LMSYS Chatbot Arena leaderboard, which ranks AI language models in a blind popularity contest.

Elon Musk's xAI Claims Its New Grok 3 AI Is Better Than ChatGPT and DeepSeek: 'Seeing the Beginnings of Creativity'

They claim Grok 3 has better accuracy, capacity, and computational power than previous models. xAI, the startup led by Elon Musk that raised $6 billion in December, has a new AI model that it claims is better than AI created by DeepSeek and ChatGPT-maker OpenAI. In a live-streamed event on X on Monday that has been viewed over six million times at the time of writing, Musk and three xAI engineers revealed Grok 3, the startup's latest AI model. They claimed Grok 3 had higher scores on math, science, and coding benchmark tests than OpenAI's GPT-4o, DeepSeek's V3, and Google's Gemini AI. Related: Elon Musk's xAI Is Reportedly Set to Hire Thousands of 'AI Tutors' With Pay Up to $65 an Hour They also said Grok 3 was a step up in sheer power from xAI's previous model Grok 2, released in August. The latest version has more than 10 times the computational power of Grok 2, greater accuracy, and a bigger capacity for large datasets. "The word Grok [means] to fully and profoundly understand something," Musk said on the livestream, noting that the word came from the 1961 novel "Stranger in a Strange Land" by American author Robert Heinlein. He added later in the livestream that "if you're using Grok 3, you may notice improvements almost every day because we're continuously improving the model." xAI engineers demonstrated how Grok 3 could be used to create code for an animated 3D plot of a spacecraft launch that started on Earth, landed on Mars, and came back to Earth. The engineers also asked Grok to combine two games, Tetris and Bejeweled, into one game. The result, which the engineers played on the livestream, was similar to Tetris with shapes inching down the screen but had the rules of Bejeweled with multicolored blocks that disappeared if there were three in a row. Related: Google's CEO Praised AI Rival DeepSeek This Week for Its 'Very Good Work.' Here's Why. Musk said that any AI could find examples of Tetris or Bejeweled online and duplicate them, but Grok 3 took it one step further. "What's interesting here is it [Grok 3] achieved a creative solution combining two games that actually works and is a good game," Musk noted. "We're seeing the beginnings of creativity." The researchers said they only trained Grok 3's reasoning abilities on math problems and competitive coding problems, but they observed that Grok 3 could apply what it learned to a variety of use cases, including reasoning through making games. xAI isn't the only major AI startup to release advanced AI this year. Last month, OpenAI released the o3-mini, its most cost-effective yet powerful model yet, while DeepSeek came out with R1, a disruptive AI model with cutting-edge performance on a less than $6 million budget.

Grok 3 wades into the AI wars

Musk's latest attempt at a 'maximally truth-seeking' bot arrives Grok 3 has begun rolling out. xAI founder Elon Musk describes the chatbot as "a maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct." Pre-training of Grok 3 was completed in early January, although the xAI team said training was ongoing. A huge datacenter was constructed for the purpose. Development began in April, and xAI said that in 122 days, the facility reached 100,000 GPUs training synchronously. It took another 92 days to expand the facility to 200,000 GPUs. It sounds like a big number, but is it really? Not when compared to the company's rivals. The rate of progress is remarkable, yet xAI sits behind OpenAI. It does, however, try to make up for this with a selection of benchmarks that put Musk's AI platform ahead of the competition. During a short presentation, the xAI team talked up Grok 3's superiority in science, mathematics, and coding when compared to other models on the market such as OpenAI's. It goes without saying that a hefty pinch of salt is required when looking at any benchmarks produced in the tech industry - not just the AI ones. xAI says that Grok 3 was developed with ten times the compute power of its predecessor. Musk claimed the figure was nearer 15 times during the presentation. The demonstrations were noteworthy, if not groundbreaking. For coding, the team asked Grok 3 to come up with a game that combined Tetris and Bejeweled, and after a short pause, the service produced a playable game that did indeed combine the best of both worlds. In a nod to Musk's Martian ambitions, the team asked for a trajectory to Mars and back, and Grok 3 produced a mission plan replete with animations. It also reported on the progress of another Musk project - SpaceX's Starship. While neither demo will set the world alight, they do highlight the breakneck pace of Grok's development. Grok 2, which debuted in 2024, will be made open source once Grok 3 is declared mature and stable. Premium+ subscribers on Musk's social media platform, X, will be the first to get their hands on the updated service. There is also an upgrade to "SuperGrok," which, according to xAI, will give early access to new features and include higher image generation limits. The price of SuperGrok was not disclosed during the presentation, although the monthly cost of a Premium+ subscription on X is $40 in the US or £17 in the UK. Musk said a conversational version was in the works and cautioned that what was being deployed this week was more a beta than anything else. "If you want a more polished version, maybe wait a week," he said. Other features that will appear in the coming weeks and months include Grok 3's models in the xAI enterprise API and the implementation of memory in conversations. Grok 3 - even with its ability to scour the internet for answers to questions - is unlikely to ruffle too many feathers, and we can well imagine competitors such as OpenAI moving swiftly to produce benchmarks demonstrating their own superiority. However, it is the pace of development that will give rivals pause for thought. Even in the fast-moving world of AI, the rapid debut of Grok 3 - though slightly later than the end-of-year timeline Musk boasted about in 2024 - along with its enhancements over its predecessor, is impressive. ®

xAI Launches Grok-3 AI Model, Claims Superior Performance Over GPT-4

Elon Musk's artificial intelligence company xAI on Tuesday announced Grok-3, and Musk is making some bold claims. The new AI model is said to have more than ten times the computing power of its predecessor and outperforms leading competitors, including OpenAI's GPT-4o and Google's Gemini. The latest iteration of xAI's flagship model introduces new "reasoning" capabilities through two distinct modes: "Think," which displays the AI's reasoning process while resolving requests, and "Big Brain" for handling more computationally intensive tasks. Alongside the model update, xAI announced Deep Search, which the company describes as a "next generation search engine." The new feature is designed to analyze information from the internet and X (Twitter) to provide comprehensive answers to user queries. Grok-3 will be available to X Premium Plus subscribers, which now costs $40 per month following a recent price increase. The company is also launching a new subscription tier called SuperGrok, priced at $30 per month, offering "the most advanced capabilities and earliest access to new features." Musk said that Grok-3 is designed to be a "maximally truth-seeking AI," even when such truth might conflict with political correctness. The model has faced previous criticism for spreading election misinformation and having fewer restrictions on text-to-image generation. Grok-3's reasoning capabilities are available in the Grok app. In the future, xAI says it plans to add synthesized voice capabilities to the Grok chatbot and intends to make the previous version, Grok-2, open source in the coming months.

How Grok 3 AI Compares to ChatGPT and Google Gemini?

The AI ecosystem has witnessed a major upheaval with the release of Grok 3 by xAI, which is geared towards taking on incumbent players such as ChatGPT and Google Gemini. Here's an exhaustive comparison that considers different aspects of these AI models, ranging from their conversational capabilities to ethics. has been built with the intent to deliver not only answers but also insightful, occasionally witty answers, similar to works such as "The Hitchhiker's Guide to the Galaxy". This gives it a clear conversational tone that might be appealing to users looking for a more engaging experience. ChatGPT, an offering, has become popular for its ability to create human-like dialogue with well-explained and context-related responses. Google's Gemini, on the other hand, leverages Google's vast knowledge base and tends to give accurate, fact-based responses with references for additional reading, making it an information powerhouse. From a user interface perspective, Grok 3 tries to reconcile the depth of conversation in ChatGPT with the accuracy of facts in Gemini, although it at times fails to deliver the integrated web browsing of Gemini.

Elon Musk's xAI to Launch Grok 3: The Next Evolution in AI Chatbots

Elon Musk's xAI Set to Challenge OpenAI with Launch of Grok 3, the "Smartest AI on Earth" Elon Musk's AI startup, , is gearing up for the launch of its next-generation chatbot, Grok 3, touted as the "Smartest AI on Earth." Scheduled for a live demo at 8:00 PM PT, Grok 3 promises to surpass current AI models, including OpenAI's ChatGPT, with its advanced reasoning and error-correction abilities. Musk recently revealed the cutting-edge capabilities of Grok 3 during his appearance at the World Government Summit in Dubai, sparking excitement among AI enthusiasts and researchers. As xAI continues to grow in prominence, its new release positions the company as a major player in the competitive AI landscape, challenging the likes of OpenAI and Google DeepMind. Musk recently spoke at the World Government Summit in Dubai, where he shared insights into Grok 3's advanced capabilities. He stated that the chatbot has powerful reasoning abilities and has demonstrated superior performance in internal testing. Unlike previous models, Grok 3 has been trained on synthetic data, enabling it to analyze its outputs, identify errors and make corrections by revisiting its dataset. This ensures logical consistency and enhances the model's overall accuracy. The in the AI sector has intensified, with xAI aiming to challenge OpenAI and other leading AI firms. Musk's departure from OpenAI in 2018 was driven by disagreements over its shift towards a for-profit structure. Recently, he has been critical of OpenAI's direction and has even filed a lawsuit against its CEO, Sam Altman, accusing the company of deviating from its original mission. Additionally, Musk and a consortium of investors made a $97.4 billion offer to acquire OpenAI's nonprofit arm. However, OpenAI has rejected the bid as Altman dismissed the proposal publicly. Beyond Grok 3, xAI has grown rapidly into one of the most valuable AI startups. It secured a in its last funding round, This was possible with backing from tech giants like Nvidia, AMD and prominent venture capital firms. The company is positioning itself as a strong contender in the AI race, alongside Microsoft-backed OpenAI and Google DeepMind. Meanwhile, the rise of has further accelerated AI development. DeepSeek's R1 model, built using older Nvidia GPUs, offers performance comparable to OpenAI's latest AI systems at a lower cost. This has brought a stiff and an aggressive competition between the U.S and China to lead in AI advancements. Those aiming to profit from AI developments should consider investing in stocks of Nvidia and Meta Platforms which demonstrate strong market potential. Analysts have rated these stocks as "Strong Buy," indicating a positive outlook for AI-driven investments. With the impending launch of Grok 3, xAI is set to make a significant impact in the AI industry. The chatbot's advanced reasoning capabilities and synthetic data training may push the boundaries of AI innovation, making it a major competitor in the market. As the AI landscape evolves, the competition between xAI, OpenAI and DeepSeek is expected to reshape the future of artificial intelligence.

What is Grok 3? Elon Musk's xAI unveils 'scary smart' AI chatbot to challenge OpenAI, DeepSeek: 10-point explainer

Elon Musk's XAI unveiled its latest artificial intelligence chatbot, Grok 3, which the tech billionaire has described as the "smartest AI on Earth". Musk's AI startup xAI launched its Grok 3 chatbot today at 9.30 am. The launch event, held on February 17, 2025, showcased the capabilities of Grok 3, including advanced reasoning, text-to-video conversion, and self-correction mechanisms. -Elon Musk praised the team behind Grok AI, saying, "Thanks to the hard work of an incredible team, and I'm honored to work with such a great team." The demo event, which was live-streamed, had nearly 100,000 viewers tuning in. -At the launch event, Musk explained the meaning of name "Grok". He said the term "grok" comes from Robert Heinlein's science fiction novel "Stranger in a Strange Land." In the novel, the word "grok" is used by a character raised on Mars and means to fully and profoundly understand something. Musk emphasised that the word conveys deep understanding and empathy, which are key attributes of Grok 3. -The Grok family of AI models are Musk's answer to foundational AI models developed by US rivals such as OpenAI's GPT-4o and Google's Gemini. They have, so far, been capable of analysing images and responding to user prompts, while also powering several generative AI features on Musk's social media platform, X. ALSO READ: JD Vance's brutal takedown of journalist for comparing White House ban to attack on free speech -Grok 3 used 100,000 Nvidia H100 GPUs to provide 200 million GPU-hours for training which exceeded Grok 2 by ten times. The large-scale installation of more computational power in Grok 3 enables it to run big datasets in a shorter time frame while providing enhanced accuracy. -Musk and xAI team claim Grok 3 is "an order of magnitude more capable" than Grok 2 released in August 2024. xAI shared benchmark comparisons, showing that Grok-3 outperformed some of the biggest names in AI, including Google's Gemini 2 Pro, DeepSeek V3, and OpenAI's GPT-4o, especially in science, coding, and math. -During the conference, Musk termed Grok 3 as "scary smart". He said: "At times I think Grok 3 is scary smart". The chatbot is expected to run on xAI's Colossus supercomputer, which reportedly uses over 100,000 Nvidia GPU hours to train AI models. The system was built in just over eight months, he added. ALSO READ: Elon Musk to unveil Grok 3 chatbot 'smartest AI on Earth' today: Check release time and key details -Beyond raw intelligence, Grok-3 flaunted a creative edge- dreaming up an entirely new game that fuses Tetris and Puyo Puyo. Grok-3 is now available to X premium+ users, the chatbot promises superior reasoning, game design capabilities, and overall performance. -Elon Musk told the crowd to update their X application to explore all the advanced features as the company released the update. He added, "We are launching a separate subscription called Super Grok for dedicated fans who want the most advanced capabilities and earliest access to new features. This is available for both the Grok app and our new website, grok.com." ALSO READ: Chase Bank to shut down all 4,700 locations across the US for 24 hours today. Here's why -"If you're looking for a more polished version, it might be worth waiting a week, but expect improvements every day. We're also working on a voice interaction feature so you can have a conversational experience. I tried it earlier today, and it's working pretty well, though it still needs some polish. The goal is to make it so you can talk to it just like you would a person. I think it's going to be one of the best experiences with Grok3, but that should be about a week away," Elon Musk added. -xAI achieved better capabilities for Grok 3 by modifying its training processes beyond hardware improvements. The updated model implements synthetic datasets, self-correction, and reinforcement learning to enhance its performance.

Elon Musk Says His New AI Chatbot Is 'Scary Smart' -- And Arriving in Weeks - Decrypt

Elon Musk announced that the next generation of his company's AI chatbot Grok may be just weeks away from release, describing it as "scary smart" and claiming it had already outperformed every other AI model in testing. The xAI CEO made these remarks during the World Governments Summit in Dubai on February 13. "At times, I think Grok-3 is kind of scary smart," Musk said. "It comes up with solutions that you wouldn't even anticipate -- you know, not obvious solutions." The chatbot developers utilized unique training methods for Grok-3. Instead of using real-world data like ChatGPT, Grok-3 relied on synthetic data and employed a self-correcting mechanism to maintain logical consistency. It got so accurate, Musk claimed, that even when it encountered incorrect information, the system reflected on the data and removed content that didn't match reality. The computational demands for training Grok-3 were massive. Experts calculate that it required 200 million GPU hours, dwarfing its Chinese competitor DeepSeek-V3's 2.7 million hours. It ran on xAI's Colossus supercluster with 100,000 Nvidia H100 GPUs -- ten times more computing power than its predecessor. Even without fine-tuning, Musk claimed the base model performed better than Grok-2. Grok-3's integration with X, Musk's social media platform, gave it the advantage of being able to scrape the social media app in real time instead of relying on browsing the web. The system can pull real-time data from X, and features what the company called "Unhinged Mode" -- which, according to xAI's own FAQ, is "intended to be objectionable, inappropriate, and offensive." The system isn't quite ready for prime time, though. Musk compared the remaining work to finishing a house: "That last 5% where you do the drywall and do the painting and the trimming -- even though it's not much work, it transforms the house." However, it may be released sooner than OpenAI's GPT-4.5, at least, which Sam Altman said could be released in weeks or months. "Probably (Grok-3) gets released in about a week or two," Elon said. He didn't clarify whether the new version would be publicly available or put behind a subscription, as happened with Grok-2 at first. Competition in the AI space has intensified. While ChatGPT dominated the market share in 2024, Chinese open-source model DeepSeek-V3 emerged as a serious contender, outperforming both GPT-4o and Meta's Llama 3.1 despite using far fewer resources. Grok was first made available on X Premium, which substantially limited its availability. It was later released free to all users of Musk's social media platform, with a new standalone website now available for everyone else. Major AI players are switching focus to reasoning models, developing AI models that are able to reflect on specific problems and find ways to solve them after a long and extensive chain of thought reasoning. The idea was first explored by Matt Schumer, back when Reflection 70b was announced. The model was trained to incorporate Chain of Thought reasoning, and was supposed to beat Claude 3.5 Sonnet at complex tasks despite being just a Llama 70b finetune. That didn't work, but just a few weeks later, OpenAI announced its "OpenAI o1" reasoning model, applying that same concept effectively. That model marked a new standard in terms of the logical capabilities AI models can exhibit, and was seen as OpenAI's moat to dominate the AI industry. But the release of DeepSeek turned the world upside down. A team of Chinese researchers built a model that was better than o1 at a fraction of the cost -- and made it open source, too. Since then, OpenAI announced that its future models would be merged into one jack-of-all-trades AI that leaves the traditional GPT architecture behind and focuses on deep reasoning first. xAI appears to be following the markets. "Grok-3 has very powerful reasoning capabilities," Elon Musk said. He didn't disclose additional information about the model's structure. The current version of Grok-2 is placed in the 18th position in the LLM Arena, well below competitors like GPT, Claude, Gemini, Qwen or DeepSeek. Looking ahead, xAI plans to scale its computing infrastructure to 1 million GPUs for future models with "trillions of parameters." The ultimate goal, according to Musk, is to advance towards artificial general intelligence through increasingly sophisticated models.

Elon Musk says Grok 3 will outperform ChatGPT, DeepSeek in the coming weeks

Elon Musk has confirmed that his AI chatbot, Grok 3 is currently being finalized and will be available in the next one to two weeks, according to Reuters. Speaking in a video call addressing the World Governments Summit in Dubai Musk described the AI tool as "scary smart." Recommended Videos "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," he said. In recent weeks several companies in the AI industry have introduced new products or updated their current tools in the wake of the Chinese startup DeepSeek unveiling its latest R1 reasoning model. Musk appears to be no different with the development of Grok 3. Prior reports have detailed that the businessman built the Colossus Supercluster, a supercomputer in Memphis Tennessee precisely for such a project. According to Business Insider, xAI, Musk's company that develops Grok plans to hire thousands of "AI tutors" this year to train the chatbot. The current staff is at approximately 900 employees. DeepSeek and its R1 model have stood out in the AI industry for being an open-source platform that is cost-effective to produce and train. Grok is notably connected to the social media platform X but can also be accessed as a standalone web-based tool, as well as via iOS and Android apps. The initial version of Grok was an open-source model. However, subsequent development of the tool has been closed-sourced and proprietary. Grok also remains limited to X users, which could prove to be a hindrance to its overall market share, Tech.co noted. Despite its popularity, DeepSeek has faced bans from several countries around the world. Additionally, several U.S. states, including Texas and New York, have banned the use of the AI tool on government-sanctioned devices. There is also a House bill being proposed that would ban the app from state-provided devices across the country. As Musk prepares to launch Grok 3, he is also embroiled in a legal case against startup, OpenAI, over whether the company should be allowed to transition from a nonprofit organization to a for-profit company. Notably one of the co-founders of OpenAI, Musk recently offered the company $97.4 billion to buy the assets of OpenAI's nonprofit. Before this, the businessman sued OpenAI in August in an attempt to halt the company's efforts to establish itself as a for-profit organization. OpenAI has stated that Musk's efforts to buy out the company and his lawsuit do not align. Having been a part of the team that initially pledged $1 billion to the development of OpenAI and its ChatGPT chatbot in 2019, Musk ultimately left when his vision for the company differed too much from the rest of the team. He wanted to integrate OpenAI into his car brand Tesla, while the rest of the team went in a different direction. Musk established the Grok chatbot in November 2023, after purchasing X, then called Twitter in 2022.

Elon Musk's xAI claims newest Grok 3 model outperforms OpenAI,...

Elon Musk's xAI claims the newest version of its flagship "Grok" chatbot outperforms rival products offered by the likes of Sam Altman-led OpenAI and China-based DeepSeek -- potentially giving the billionaire an edge in the AI arms race. Dubbed "Grok 3," Musk's new AI model scored higher on tests in math, science and coding than OpenAI's GPT-4o, Google's Gemini, Anthropic's Claude and DeepSeek's V3 models, according to a chart released by the startup during a live-streamed launch event late Monday. Musk said Grok 3 would be "an order of magnitude more capable" than its previous version "in a very short period of time." Grok 3 used 10 times of the computing power than Grok 2 during its development. "Grok-3 across the board is in a league of its own," Musk added. The chatbot, which was previously described by Musk as "scary smart," is available to premium subscribers on X, formerly known as Twitter. Musk's startup also revealed a new tool called "Deep Search," an AI search engine powered by Grok that explains the reasoning behind its responses to user queries. The claims about Grok 3's performance have yet to be independently verified. Andrej Karpathy, an OpenAI co-founder and former director of AI at Tesla, said after his initial testing that Grok 3 "clearly has an around state of the art thinking model" and described it as "slightly better" than recent releases by DeepSeek and Google. Launched in 2023, xAI is in talks to raise $10 billion at a whopping $75 billion valuation, with investment giants such as Andreessen Horowitz and Sequoia Capital slated to participate. Grok 3 launched in the midst of a legal slugfest between Musk and Altman over the future of OpenAI. Musk co-founded the ChatGPT maker in 2015, but left the firm after disagreements with Altman over its long-term direction. Musk has a pending federal antitrust lawsuit against OpenAI and its key investor Microsoft. He is also seeking a preliminary injunction to block Altman's plans to transform OpenAI from a nonprofit to for-profit entity. Last week, Altman and his allies flatly rejected Musk's unsolicited $97.4 billion offer to take control of OpenAI. Musk had said in a court filing that he would abandon the hostile takeover effort if Altman dropped his plans to become a for-profit. Elsewhere, DeepSeek caused shockwaves throughout the US tech sector last month after it released an open-source chatbot that was on par with US rivals. DeepSeek claimed that the model cost less than $6 million to train and that it was developed without access to Nvidia's most powerful computer chips, which are subject to US export controls and considered necessary to power advanced AI models. Some experts, including Musk, have expressed doubt about DeepSeek's claims and asserted that the Chinese firm likely has far more chips than it has publicly admitted.

Elon Musk's xAI Unveils Grok 3 Family of AI Models With These New Features

Grok 3 family of artificial intelligence (AI) models were announced on Monday. The successor to the Grok 2 models was announced in a live event hosted by xAI engineers and the company owner Elon Musk. The new series contains several large language models (LLMs) with different parameter sizes and reasoning-based variants. The new models will come with new features such as DeepSearch and a voice mode. Alongside, the company also announced a new subscription tier dubbed SuperGrok to access higher rate limits and certain new features. Musk hosted a live stream on X (formerly known as Twitter) showcasing the capabilities of the new AI models. Calling Grok 3 "an order of magnitude more capable than Grok 2," the billionaire highlighted that xAI built its new data centre to pre-train the LLMs, and the first 100,000 GPUs were running within 122 days. The capacity was further doubled in the next 92 days. The Grok 3 family comprises several LLMs, but not all of them are available currently. At the event, the company unveiled Grok 3, Grok 3 mini (a smaller but faster model), alongside the Grok 3 Reasoning and Grok 3 mini Reasoning (test time compute-based reasoning models). The rollout for the announced AI models has begun, but some other models are currently in the beta phase. Musk highlighted that some of the chain-of-thought (CoT) steps will be obscured in the Grok 3 reasoning models to prevent instances of distillation. Notably, distillation is the process when the synthetic data generated from one AI model is used to train another, smaller model. Two new features were also unveiled with the Grok 3 family. First is DeepSearch, which is xAI's version of the Deep Research feature recently launched by OpenAI and Google. It scours the Internet and the X platform to analyse information when a complex query is asked and generates a comprehensive report. Grok 3 will also get a voice mode that will allow the AI model to respond to queries verbally. However, this feature will not be available at launch. Musk said it could be rolled out "as soon as a week from now." Coming to benchmarks, the company claimed that Grok 3 outperforms GPT-4o on the American Invitational Mathematics Examination (AIME), Graduate-Level Google-Proof Q&A (GPQA), and LiveCodeBench benchmarks. It is also claimed to achieve higher scores than the Claude 3.5 Sonnet, DeepSeek-V3, and Gemini-2 Pro based on internal testing. xAI also said the Grok 3 reasoning models outperform OpenAI's o3 mini model. Alongside the new models, xAI also introduced the new SuperGrok subscription tier. While pricing details were not revealed, this tier will offer features such as the DeepSearch and Think (reasoning) mode, higher image generation limits, and early access to new features. Those subscribed to X Premium Plus will get access to Grok 3, however, other features are not available at this tier. Musk also said that the company will adopt the policy of open-sourcing the last version of the AI model once the current version is fully rolled out. He added that once Grok 3 is stable and mature, which could take up to a few months, Grok 2 will be released in open-source.

Grok 3: Elon Musk's new xAI chatbot in the market

In a live stream presentation with xAI engineers, Musk claimed that Grok 3 has "over 10 times" the computing power of its predecessor, Grok 2.Grok 3, "the smartest AI on Earth" as Elon Musk describes it, launched today. Musk, in a live stream presentation with xAI engineers, claimed that Grok 3 has "more than 10 times" the compute power of its predecessor Grok 2. Musk's claims Musk claimed that Grok-3 outperforms Alphabet Inc.'s Google Gemini, DeepSeek's V3 model, Anthropic's Claude, and OpenAI's GPT-4o across mathematics, science, and coding benchmarks. During the livestream he said, ""People are going to fall in love with Grok. That's 1,000% probable." Musk revealed that the model finished pre-training in early January and is already making daily improvements. "Literally within 24 hours, you'll see improvements," he said. Discussing the goals of xAI and Grok, Musk stated that their aim is to "understand the universe." "We want to answer the biggest questions: Where are the aliens? What's the meaning of life? How does the universe end?" He added that in order to understand the nature of the universe we "must rigorously pursue truth" even when the "truth" is at odds with what is "politically correct". He also didn't forget to praise the team behind Grok AI, saying, "Thanks to the hard work of an incredible team, and I'm honoured to work with such a great team." Access Initially, access to this advanced AI will be exclusive to Premium Plus subscribers on the X platform, with a new subscription tier, "Super Grok," providing enthusiasts with early and comprehensive access to its features. Grok 2 Grok 2 was launched in August last year. It boasted of salient features like recently introduced image generator Aurora and web search among others. The Grok AI chatbot is available to anyone for free by signing up on X. Funding Musk's startup xAI secured $6 billion in its latest funding round in December. The company received backing from US venture capitalists, chipmakers Nvidia and AMD, as well as investment funds from Saudi Arabia, Qatar, and others. The company is now valued at $50 billion, making it one of the world's most valuable startups, though still far behind its competitor OpenAI's $157 billion valuation. Musk vs OpenAI Last week, a group of investors led by Musk offered $97.4 billion to acquire OpenAI's non-profit assets. OpenAI had stated its goal to become a for-profit organisation to secure the funding needed for developing top AI models. But the deal was rejected by the board of directors of OpenAI. The dispute between Musk and OpenAI began when he filed a lawsuit last year alleging that OpenAI has strayed from its original mission as a non-profit research lab meant to serve the public good. Musk cofounded OpenAI with its CEO Sam Altman in 2015 to develop AI technology in a way that "benefited humanity".

Elon Musk Unveils Latest Grok AI Model, Claims to Outperform DeepSeek, OpenAI

Elon Musk's artificial-intelligence startup, xAI, has unveiled its latest AI model, Grok 3, claiming it outperforms DeepSeek and OpenAI models across various benchmarks. Musk said the latest version has more than 10 times the computing power of its previous version, speaking in a livestream on Monday, alongside three members of xAI's engineering team. Grok 3 completed its pre-training in early January, they added. Citing its own comparisons using math, science and coding benchmarks, xAI claimed Grok 3's outperforms Google's Gemini 2.0 Pro, DeepSeek's V3 model, Anthropic's Claude 3.5 Sonnet and OpenAI's GPT-4o. "We should emphasize that this is kind of a beta-like, meaning that you should expect some imperfections at first, but we will improve it rapidly, almost every day," Musk said. The xAI team added that it built its own data "to build the best AI out there," utilizing 200,000 graphics processing units. The company first launched its stand-alone consumer chatbot, Grok, in January, which shares its name with the AI language model. Musk added that a voice function to interact with the chatbot is "about a week away" from being publicly released. In addition, xAI introduced a new search engine product called DeepSearch, which it described it as the "first generation of our Grok agents." A demonstration showed a chatbot-like interface that can conduct extensive research, analyze data and assist with coding. DeepSearch can also explain its reasoning process, providing insight into how it answers questions and plans responses. Grok 3 features and the DeepSearch function will be first available to Premium+ subscribers on the social media platform X, with a broader rollout planned in the coming days. It also introduced a new subscription plan called SuperGrok for its mobile app and website, offering subscribers access to the latest Grok features and updates. The latest announcement comes amid a frenzy triggered by Chinese AI startup DeepSeek, after launching its open-source AI model, R1. DeepSeek claims the model excels at problem solving, rivaling OpenAI's GPT-4o reasoning model and most notably, at a fraction of the cost per use. This has raised concerns over the demand for advanced chips and data centers, while also boosting optimism in China's AI sector. In recent weeks, Chinese e-commerce giant Alibaba released its upgraded AI model, while tech giant Baidu said it will make its AI chatbot available to use free-of-charge, adding to the growing competition.

Elon Musk's xAI unveils Grok-3 with advanced reasoning capabilities - SiliconANGLE

Elon Musk's xAI unveils Grok-3 with advanced reasoning capabilities Elon Musk's xAI Corp. late Monday night announced the launch of Grok-3, the latest in the company's family of large language models. The company says the AI model is a significant leap in power over its previous Grok-2 LLM that includes "reasoning models" to mimic human thinking. During a livestream announcement, xAI researchers said that Grok-3 was trained using 10 to 15 times more compute power than Grok-2. The company launched a massive supercomputer training system named Colossus in September, built with 100,000 Nvidia Corp. H100 graphics processing units. It's designed to bring new iterations of Grok online. "Grok-3 across the board is in a league of its own," Chief Executive Elon Musk said during the announcement, claiming that the model can outperform models from OpenAI and China's DeepSeek based on early testing in math, science and coding. Grok-3 includes two main reasoning models: Grok-3 Reasoning beta, a large complex model, and Grok-3 mini Reasoning, a small fast model capable of generating quick answers. With the models enabled on xAI's chatbot, the models will reveal their "thinking" as they do step-by-step reasoning during complex science, mathematics and programming questions. This release comes at a time when other companies have started to release reasoning models that break down complex tasks into smaller tasks and then attempt to fact-check themselves before offering solutions. The objective is to provide a better result. Example models include those developed by competing companies such as OpenAI's o1 and o3-mini reasoning, DeepSeek's R1 and Google LLC's Gemini 2.0 Flash Thinking Experimental. "We should emphasize that this is kind of a beta, meaning that you should expect some imperfections at first, but we will improve it rapidly, almost every day," Musk added. To access the reasoning capabilities of the Grok-3 models, users can turn on "Think" to have it reason through their queries. And for more difficult questions, they can activate "Big Brain" mode, which xAI says is best suited for more complicated queries that involve reasoning in math, science or coding. The reasoning mode for Grok-3 can also be paired with a search capability called "DeepSearch," which takes longer but causes the model to scan the internet for relevant knowledge and incorporate it into its answers. XAI said using DeepSearch will result in more relevant, detailed responses. The addition of this deep internet research capability will mean that xAI's models will join rival models which have similar features including OpenAI and Google. Perplexity AI Inc., the creator of an AI search engine, bakes deep internet research directly into its service when presenting search answers. Grok-3 will also receive a voice mode that will allow it to respond to queries verbally. Although the feature will not be available at launch, Musk said it will be rolled out in a week or so. The voice mode for Grok will be more than voice-to-text, Musk added. It will understand tone, inflection and pacing, and "it will be like talking to a person."

Neural: xAI unveiling Grok 3 tonight -- could GPT-4.5 steal the show? - 9to5Mac

Welcome to Neural. AI moves fast. We help you keep up. OpenAI says GPT-4.5 is coming to ChatGPT in a matter of weeks. But first, xAI will unveil Grok 3 tonight. Now the countdown is on for OpenAI to do the funniest thing ever... Fresh off the heels of threatening to buy OpenAI, Elon Musk announced on Saturday that Grok 3 will arrive tonight in the form of a live demo scheduled for 8 p.m. PT. Musk hypes up Grok 3 by calling it the "smartest AI on Earth." Actual details around Grok 3 remain a mystery until tonight's reveal. My biggest question about Grok: When will it come to Tesla? AI-driven (supervised) self-driving is one thing, but drivers could benefit from a Grok-powered voice assistant that replaces the basic voice control function. Question number two is whether or not OpenAI will steal xAI's thunder tonight and do something fun with GPT-4.5. OpenAI's Sam Altman laid out the roadmap for ChatGPT last week. Altman said GPT-4.5 is coming in a matter of weeks, which suggests a release today is unlikely, followed by GPT-5 in a matter of months. GPT-5 is the more ambitious model that aims to unify fast large language models with slower thinking reasoning models -- no need to choose between a half dozen models per query. Meanwhile, Altman continues to vaguely tease the benefits of GPT-4.5 on X: When asked to "steal the show" tonight, Altman responded diplomatically with "that wouldn't be very nice..." Separately, OpenAI has deployed some actual releases over the last few days: Then there was this all-time winner for most vague hype post ever: The update, apparently, is meant to make GPT-4o a better writer, especially when given examples to follow, and less of an AI-slop generator. If nothing else, the last week in OpenAI model updates shows just how critical GPT-5 will be for ChatGPT. The flow chart for which models are best at what and how they're limited is getting beyond unwieldy, and it's outdated before anyone can actually put such a chart together. Perhaps the most intriguing news of all in the last few days is that Anthropic plans to release Claude 4 soon. Like GPT-5, Claude 4 is expected to be a single chatbot that includes quick responses like Claud 3.5 while also handing reasoning queries that are more resource-intensive.

xAI's Grok 3 to be Released Today

Besides. xAI is in talks for a $10 billion funding, and a $5 billon plus deal to purchase servers from Dell. Elon Musk, CEO of xAI, revealed that the company will release its latest AI model, Grok 3, on Monday. Musk calls it the smartest AI on earth. Users across social media are anticipating Grok 3's capabilities, given that it has been trained on 100,000 NVIDIA GPUs, ten times more computing power than its predecessor. In addition to the release of Grok 3, xAI is reportedly in talks to raise $10 billion in funding. Bloomberg reported this on Saturday and said Sequoia Capital, Andreessen Horowitz, and Valor Equity Partners will participate in the round. This will increase xAI's valuation to $75 billion. In December last year, the company completed a $6 billion Series C funding round, with participation from firms such as Andreessen Horowitz, BlackRock, Sequoia Capital, NVIDIA, and AMD. xAI is currently valued at $51 billion. Furthermore, Bloomberg reported on Friday that xAI is in the 'advanced stages' of securing a deal worth over $5 billion with Dell Technologies for servers. Dell will reportedly sell xAI servers containing NVIDIA GB200 this year to handle AI workloads. Moreover, xAI's GPU cluster Colossus is claimed by the company to be the world's most powerful AI training system. Musk notes that this cluster was built in 122 days from start to finish. According to an earlier report from Financial Times, xAI plans to expand Colossus by over ten times to include more than 1 million GPUs. However, given how DeepSeek disrupted the market by building an AI model with significantly fewer GPUs and beating most of the competition, NVIDIA lost over $500 billion in its market cap as shareholders questioned the demand for AI hardware. Grok 3 might answer the question of how much an AI model can improve by incorporating exponentially higher amounts of computing power, contributing new evidence to the ongoing debate about AI scaling laws. When AIM reached out to Paras Chopra, founder of Lossfunk, he said, "Performance is often log-linear. So I'd say 10x more compute would have a ~double jump in performance over Grok 2." Currently, Grok AI is available for free as a chatbot on X.com. It will have to compete not only with the existing models but also with the upcoming hybrid model by Anthropic and GPT 4.5, which is part of OpenAI's newly released roadmap.

Musk's xAI releases artificial intelligence model Grok 3, claims better performance than rivals in early testing

Grok 3 features will be rolled out for premium X members starting today, while the model will also be accessible through a separate subscription for the Grok web and app version. Elon Musk's xAI on Tuesday unveiled its latest artificial intelligence model, Grok 3, claiming it can outperform offerings from OpenAI and China's DeepSeek based on early testing, which included standardized tests on math, science and coding. "We're very excited to present Grok 3, which is, we think, an order of magnitude more capable than Grok 2 in a very short period of time," Musk said at a demonstration of Grok 3 that was streamed on his social media platform X. The team also said it was launching a new product called "Deep Search," which would act as a "next generation search engine." Grok 3 will be rolled out for premium X subscribers later in the day, and will also be accessible through a separate subscription for the model's web and app versions, the xAI team said. Speaking at The World Governments Summit in Dubai last week Musk had dubbed the model "scary smart," with powerful reasoning capabilities, claiming it outperformed all other existing models in xAI's internal tests. "This might be the last time that an AI is better than Grok," Musk said at the time, adding that it was trained on "a lot of synthetic data," and was capable of reflecting upon its mistakes to achieve logical consistency. The xAI team claimed that an early iteration of Grok 3 had been given better ratings than existing competitors on Chatbot Arena, a crowdsourced website that pits different AI models against each other in blind tests. Toward the end of the product demo, Musk said that the company will keep improving the model. "We should emphasize that this is kind of a beta, meaning that you should expect some imperfections at first, but we will improve it rapidly, almost every day," he said, adding that the voice assistance for the model would be released at a later time.

Elon Musk's xAI adds 'Big Brain' reasoning to Grok-3

Jess Weatherbed is a news writer focused on creative industries, computing, and internet culture. Jess started her career at TechRadar, covering news and hardware reviews. Elon Musk's xAI unveiled Grok-3 on Tuesday, announcing that the new artificial intelligence model has "more than 10 times" the compute power of its predecessor. xAI said its latest flagship outperforms OpenAI's GPT-4o, Google's Gemini, and DeepSeek's V3 models in early testing, and now features "advanced reasoning" capabilities. So-called reasoning models are trained to answer more complex questions by breaking instructions down into smaller tasks and attempting to fact-check themselves before offering a solution, with the aim of providing stronger results. Similar models have been developed by rival companies, including OpenAI's o1, DeepSeek's R1, and Google's Gemini Flash Thinking. There are two Grok-3 reasoning modes available: "Think", which will display Grok's reasoning as it resolves requests; and "Big Brain" for complex tasks that require more computational power. xAI is also launching a Grok AI agent product called Deep Search, which the company describes as a "next generation search engine." Musk says that Grok-3 is a "maximally truth-seeking AI -- even if that truth is sometimes at odds with what is politically correct." Previous versions of the xAI chatbot have been criticized for spreading election misinformation and having fewer guardrails on text-to-image generation, allowing it to spit out questionable or offensive imagery. OpenAI is also exploring how to develop its models to "seek the truth" when handling controversial topics, but with the aim of maintaining certain safety rails. The Grok-3 reasoning capabilities are available in the Grok app for subscribers to X Premium Plus, which now starts at $40 per month. This is the second hike for Premium Plus in two months, having increased from $16 to $22 in December. xAI said it is also launching a new subscription plan called SuperGrok that will provide "the most advanced capabilities and earliest access to new features." SuperGrok will reportedly cost $30 per month, though it's unclear if this is an additional charge on top of X subscriptions. Elon Musk said that the Grok chatbot will soon gain a synthesized voice feature that sounds similar to OpenAI's Advanced Voice Mode for ChatGPT. xAI is also planning to make Grok-2 open source in the coming months.

9to5Neural: xAI unveiling Grok 3 tonight -- could GPT-4.5 steal the show? - 9to5Mac

Welcome to 9to5Neural. AI moves fast. We help you keep up. OpenAI says GPT-4.5 is coming to ChatGPT in a matter of weeks. But first, xAI will unveil Grok 3 tonight. Now the countdown is on for OpenAI to do the funniest thing ever... Fresh off the heels of threatening to buy OpenAI, Elon Musk announced on Saturday that Grok 3 will arrive tonight in the form of a live demo scheduled for 8 p.m. PT. Musk hypes up Grok 3 by calling it the "smartest AI on Earth." Actual details around Grok 3 remain a mystery until tonight's reveal. My biggest question about Grok: When will it come to Tesla? AI-driven (supervised) self-driving is one thing, but drivers could benefit from a Grok-powered voice assistant that replaces the basic voice control function. Question number two is whether or not OpenAI will steal xAI's thunder tonight and do something fun with GPT-4.5. OpenAI's Sam Altman laid out the roadmap for ChatGPT last week. Altman said GPT-4.5 is coming in a matter of weeks, which suggests a release today is unlikely, followed by GPT-5 in a matter of months. GPT-5 is the more ambitious model that aims to unify fast large language models with slower thinking reasoning models -- no need to choose between a half dozen models per query. Meanwhile, Altman continues to vaguely tease the benefits of GPT-4.5 on X: When asked to "steal the show" tonight, Altman responded diplomatically with "that wouldn't be very nice..." Separately, OpenAI has deployed some actual releases over the last few days: Then there was this all-time winner for most vague hype post ever: The update, apparently, is meant to make GPT-4o a better writer, especially when given examples to follow, and less of an AI-slop generator. If nothing else, the last week in OpenAI model updates shows just how critical GPT-5 will be for ChatGPT. The flow chart for which models are best at what and how they're limited is getting beyond unwieldy, and it's outdated before anyone can actually put such a chart together. Perhaps the most intriguing news of all in the last few days is that Anthropic plans to release Claude 4 soon. Like GPT-5, Claude 4 is expected to be a single chatbot that includes quick responses like Claud 3.5 while also handing reasoning queries that are more resource-intensive.

Musk's xAI unveils Grok-3 AI chatbot to rival ChatGPT

Elon Musk's artificial intelligence startup xAI has introduced Grok-3, the latest iteration of its chatbot, as it looks to compete with Chinese AI firm DeepSeek, Microsoft-backed OpenAI, and Alphabet's Google. Grok-3 debut comes at a critical moment in the AI arms race, just days after DeepSeek unveiled its powerful open-source model and as Musk moves aggressively to expand xAI's influence. The chatbot is being rolled out immediately to Premium+ subscribers on X, the social media platform owned by Musk. xAI is also launching a new subscription tier, SuperGrok, for users accessing the chatbot via its mobile app and Grok.com website. "Grok-3 across the board is in a league of its own," Musk said during a livestream alongside three xAI engineers late on Monday, adding that the model significantly outperforms its predecessor, Grok-2. Last week, a consortium of investors led by Musk offered US$97.4 billion to acquire OpenAI's nonprofit assets, an offer the ChatGPT-maker rejected. Musk on Monday reiterated xAI's commitment to open-source AI, saying earlier versions of Grok will be made publicly available once the latest model reaches full maturity. He expects Grok-3 to meet that benchmark in a few months. The latest release introduces a smart search engine, called DeepSearch, which xAI describes as a reasoning-based chatbot capable of articulating its thought process when responding to user queries. The tool, demonstrated during the livestream, offers functions for research, brainstorming, and data analysis. As competition in AI intensifies, xAI is ramping up its data center capacity to train more advanced models. Bloomberg News reported last week the startup is in discussions to raise up to $10 billion in funding, which could value the company at around $75 billion.

Elon Musk Reveals Grok 3 AI Chatbot: Here's What It Can Do

Elon Musk's xAI on Monday unveiled its latest flagship model, Grok 3, alongside what appears to be a big price hike for X Premium+ subscriptions. Grok 3, which Musk has dubbed the "smartest AI on Earth," is first rolling out to Premium+ subscribers of the X app. However, it will soon be available via a "Super Grok" subscription on the Grok app and Grok.com website. The latest large language model (LLM) from xAI has been trained on 200,000 GPUs and uses more than 10x the computing power of Grok 2. It has advanced reasoning and agentic abilities and beats almost all of its rivals on math, science, and coding benchmarks. To access Grok 3's reasoning abilities, users can click on the new "Think" and "Big Brain" buttons. "Think" relies on a smaller Grok 3 mini model and can solve simple queries, whereas "Big Brain" relies on Grok 3 and can be used to solve more complex queries. In a demo, an xAI engineer used Big Brain to create a game that combines Tetris and Bejeweled. When users drop prompts via Think or Big Brain, Grok 3 and Grok 3 mini will display their "thoughts" on screen. To view their approach to generating a response, users can maximize the processing window. However, not all of Grok's thoughts will be revealed, Musk said. Some of the thoughts will be obscured to stop rival companies from copying it, he added. On a mathematics benchmark called AIME 2025, Grok 3 and Grok 3 mini outperformed OpenAI's 03 mini, DeepSeek's R1, and Google's Gemini 2 Flash models in reasoning abilities, xAI says. In addition to reasoning, Grok 3 also brings a new agentic feature called DeepSearch. This feature can be used to conduct a comprehensive analysis and generate a report. When prompted, it opens a new progress bar on the left and displays Grok's thoughts on the right. The right-hand panel also shows the websites Grok accessed and presents the final output with key citations. OpenAI, Google, and Perplexity recently released similar agentic research tools; all of them are called "Deep Research." Some Grok 3 features and models are launching in beta, TechCrunch notes. Voice mode has been pushed by a week since it's "still a little patchy," Musk said in an X post. Premium+ subscribers should see Grok 3 appear on X. If not, update your app. Those subscribers, however, may be in for a big price hike. As TechCrunch notes, the Premium support page lists a Premium+ subscription at $40 per month, up from $22. On the sign-up page, you can get a 17% discount right now, so a monthly subscription is just under $33. This is the second hike for the Premium+ tier in three months. In December, X increased the cost by 37.5%. Premium+ is the most expensive plan in X's subscription model. It offers an ad-free experience, the ability to publish articles, access to all the features of Grok AI, and the highest priority for comments, among other perks.

Elon Musk's Grok 3 Launch: Know How its Deep Search & Advance Reasoning Work

Elon Musk's Grok 3: Revolutionary AI with Deep Search and Voice Interaction In a notable move, xAI launched Grok 3 on February 18, and Elon Musk is already claiming it to be the leading advancement in artificial intelligence. Based on recent announcements and demonstrations, Grok 3 is claimed to have outperformed both ChatGPT and DeepSeek in various benchmarks, particularly in areas like math, science, and coding. According to Musk, demonstrates "scary smart" capabilities and exceeds current AI models in performing various types of tasks. xAI has made substantial advancements while establishing itself as a major player in the competitive AI market. However, these claims are primarily based on internal metrics and demonstrations by xAI, and there's some skepticism about these assertions without independent evaluations.

Musk's xAI unveils Grok-3 AI chatbot to rival ChatGPT, China's DeepSeek

(Reuters) - Elon Musk's artificial intelligence startup xAI has introduced Grok-3, the latest iteration of its chatbot, as it looks to compete with Chinese AI firm DeepSeek, Microsoft-backed OpenAI, and Alphabet's Google. Grok-3 debut comes at a critical moment in the AI arms race, just days after DeepSeek unveiled its powerful open-source model and as Musk moves aggressively to expand xAI's influence. The chatbot is being rolled out immediately to Premium+ subscribers on X, the social media platform owned by Musk. xAI is also launching a new subscription tier, SuperGrok, for users accessing the chatbot via its mobile app and Grok.com website. "Grok-3 across the board is in a league of its own," Musk said during a livestream alongside three xAI engineers late on Monday, adding that the model significantly outperforms its predecessor, Grok-2. Last week, a consortium of investors led by Musk offered$97.4 billion to acquire OpenAI's nonprofit assets, an offer the ChatGPT-maker rejected. Musk on Monday reiterated xAI's commitment to open-source AI, saying earlier versions of Grok will be made publicly available once the latest model reaches full maturity. He expects Grok-3 to meet that benchmark in a few months. The latest release introduces a smart search engine, called DeepSearch, which xAI describes as a reasoning-based chatbot capable of articulating its thought process when responding to user queries. The tool, demonstrated during the livestream, offers functions for research, brainstorming, and data analysis. As competition in AI intensifies, xAI is ramping up its data center capacity to train more advanced models. Bloomberg News reported last week the startup is in discussions to raise up to $10 billion in funding, which could value the company at around $75 billion. (Reporting by Surbhi Misra in Bengaluru; Editing by Arun Koyyur)

Musk says Grok 3 is "scary smart" -- but how does it work?

Elon Musk has announced the upcoming release of Grok 3, the latest version of his artificial intelligence chatbot, scheduled for Monday night at 8:00 p.m. Pacific Time. A live demo will accompany the launch, which will likely take place on the platform X, formerly known as Twitter. Musk labeled Grok 3 as "the smartest AI on Earth" and indicated that he would be logging offline until the demo to assist the team in preparation for the launch. Musk previously indicated that Grok 3 will incorporate additional datasets into its training set, including "all court cases," which he claims will produce "extremely compelling legal verdicts." In recent weeks, he has shared examples demonstrating Grok 3's capabilities, including references to "Testing Grok 3 int4 inference." xAI first released a beta version of Grok 2 in August, followed by the launch of the company's first public image generation in December. In the same month, xAI made Grok freely available for all X users, expanding access beyond the Premium Plus subscribers who could use it previously. The company is also reportedly developing a dedicated app for Grok, allowing users to access the AI model without needing to use X. iPhone users can now access Grok on the App Store Founded in July 2023, xAI was established by Musk following his departure from OpenAI, the AI firm he co-founded in 2015 alongside Sam Altman. The imminent launch of Grok 3 comes shortly after Musk led a group of investors in a $97.4 billion bid for control of OpenAI. Musk has criticized Altman and OpenAI's transition from a non-profit to a for-profit model, accusing Altman of being a "Swindler" and "Scam Altman" in response to the recent developments. During a video conference at the World Government Summit in Dubai, Musk mentioned that "at times I think Grok 3 is scary smart." He explained that the model was trained on synthetic data and has the capability of reflecting on its mistakes, allowing it to achieve logical consistency by reviewing the data iteratively. This announcement occurs amidst heightened competition in the AI landscape, as various entities, including China's DeepSeek, strive to create competitive models.

Musk debuts Grok-3 AI chatbot to rival OpenAI, DeepSeek

Elon Musk's artificial intelligence startup xAI debuted its updated Grok-3 model, showcasing a version of the chatbot technology to challenge OpenAI days after the billionaire's unsolicited cash bid to buy the company was rejected. Across math, science and coding benchmarks, Grok-3 beats OpenAI's GPT-4o, Alphabet's Google Gemini, DeepSeek's V3 model and Anthropic's Claude, xAI said via a livestream on Monday. Grok-3 has "more than 10 times" the compute power of its predecessor and completed pre-training in early January, Musk said in a presentation alongside three xAI engineers. Musk's performance claims, which have not been independently verified, ramp up an increasingly bitter rivalry between his startup and OpenAI. He launched xAI in 2023 as an alternative to the ChatGPT maker, which he's publicly criticized for its plans to restructure as a for-profit business. Musk, the world's richest person, has filed two lawsuits against OpenAI for allegedly straying from its founding principles and offered to buy OpenAI's nonprofit arm for $97.4 billion in a bid that was rejected last week. OpenAI Chief Executive Officer Sam Altman classified the bid as a tactic to "slow us down." Musk was involved in OpenAI's founding but has been critical of the company since leaving the board in 2018. XAI also introduced a new smart search engine with Grok-3, calling it DeepSearch. DeepSearch is a reasoning chatbot that expresses its process of understanding a query and how it plans its response. It includes options for research, brainstorming and data analysis, the demonstration showed. Musk's team also said it intends to release a voice-based chatbot. "We're continually improving the models every day, and literally within 24 hours, you'll see improvements," Musk said. Grok-3 is available Premium+ subscribers on X, a service that costs $22 a month. That compares to $200 a month for full access to OpenAI's GPT-4o. xAI is starting a new subscription called SuperGrok for the bot's mobile app and Grok.com website, and plans to open-source preceding versions of Grok models as soon as the latest one is fully mature. Musk said he expects that transition to be complete for Grok-3 in a few months. After the Grok-3 updates were released, Andrej Karpathy, an OpenAI co-founder no longer at the company, posted a preliminary review of the new model on X, writing that it "feels somewhere around the state of the art territory of OpenAI's strongest models." But the computer scientist, who formerly lead AI efforts at Tesla Inc., said Musk's model also fabricated facts and lagged behind in certain functions. Karpathy said more evaluations are needed over the next days and weeks to get a better idea of the model's capabilities. AI powerhouses like OpenAI and xAI have raised funds at a rapid clip with valuations soaring. Musk's xAI is in talks to raise about $10 billion in a funding round that would value the company at roughly $75 billion, Bloomberg News reported last week. The company was last valued at about $51 billion, according to data compiled by PitchBook. OpenAI is in talks to raise as much as $40 billion in a round that would push its valuation to as much as $300 billion. These businesses are also capital-intensive. SoftBank Group Corp., OpenAI, Oracle Corp. and Abu Dhabi-backed MGX jointly announced a program in January to deploy $100 billion, with the goal of eventually spending $500 billion, for the construction of data centers and other infrastructure for AI in the US. Dell Technologies Inc. is at an advanced stage of securing a deal worth more than $5 billion to provide xAI with servers optimized for AI. But rival technologies are emerging that could challenge this model and make it easier for new competitors to emerge. Last month, Chinese AI company DeepSeek released a new open-source AI model, called R1, that matched or beat leading US competitors on a range of industry benchmarks. The company said it built the model for a fraction of the cost of its US counterparts.

Musk Debuts Grok-3 AI Chatbot to Rival OpenAI, DeepSeek

Elon Musk's artificial intelligence startup xAI showed off the updated Grok-3 model, showcasing a version of the chatbot technology that the billionaire has said is the "smartest AI on Earth." Across math, science and coding benchmarks, Grok-3 beats Alphabet Inc.'s Google Gemini, DeepSeek's V3 model, Anthropic's Claude and OpenAI's GPT-4o, the company said via a live stream on Monday. Grok-3 has "more than 10 times" the compute power of its predecessor and completed pre-training in early January, Musk said in a presentation alongside three of xAI's engineers.

Musk debuts Grok-3 AI chatbot to rival OpenAI, DeepSeek

Elon Musk's artificial intelligence startup, xAI, showed off the updated Grok-3 model, showcasing a version of the chatbot technology that the billionaire has said is the "smartest AI on Earth." Across math, science and coding benchmarks, Grok-3 beats Alphabet's Google Gemini, DeepSeek's V3 model, Anthropic's Claude and OpenAI's GPT-4o, the company said via a live stream on Monday. Grok-3 has "more than 10 times" the computing power of its predecessor and completed pretraining in early January, Musk said in a presentation alongside three of xAI's engineers. "We're continually improving the models every day, and literally within 24 hours, you'll see improvements," Musk said.

Musk claims Grok 3 is 'outperforming' rivals, with full release imminent

We'll soon get the third version of the first AI 'with a sense of humor.' Elon Musk has previously confirmed that Grok 3 will arrive soon. This AI model was trained at the Colossus Supercluster with its 100,000 GPUs. However, the billionaire narrowed down the timescale at the World Government Summit in Dubai, suggesting Grok 3 will be ready in two or three weeks, reports Reuters. Musk also claimed that Grok 3 will be more powerful than any AI out there, including the groundbreaking DeepSeek AI. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok is outperforming anything that's been released, that we're aware of, so that's a good sign," said Musk during a video call where he was addressing the delegates at the international summit. He also said that it's in the final stages of development and that it will be released in a week or two, reports the source. When Grok 3 finally arrives, we can compare it with other existing AI LLMs and see how effective Musk's investment in the Colossus Supercluster is. After all, he had to beg for these GPUs while having dinner with Nvidia CEO Jensen Huang, so we're interested if his begging and billions paid off. This is the second time that a member of the Trump administration has appeared at an international summit involving world leaders, with VP JD Vance proclaiming at the Paris AI Action Summit that the most powerful AI chips will be built in America. While it's unclear if Musk appeared on screen as a representative of the White House or not, he is one of the heads of the Subcommittee on Delivering on Government Efficiency, also known as DOGE, which is tasked with streamlining the government and slashing federal spending. In related news, Musk has made a $100-billion offer for OpenAI as the ChatGPT maker is seeking to transition from being a non-profit to a for-profit organization. The latter is quick to reject this bid, though, especially as Musk has previously sued it to block the move. OpenAI CEO Sam Altman has said that this move was Musk's attempt to destabilize the company; the company also said that this offer clashes with Musk's previous lawsuit. While Musk is leading a takeover attempt of OpenAI, his own AI company, xAI, is currently pushing for more funding. It has recently secured $6 billion, doubling its total capital raised and putting its valuation at $50 billion. This amount should be enough to secure 100,000 Nvidia GPUs, which is precisely what he plans for the Colossus Supercluster in Memphis.

Elon Musk's xAI unveils Grok 3 model to take on ChatGPT and Google Gemini | BreakingNews.ie

Elon Musk's xAI has unveiled its latest flagship AI model, Grok 3, which it claims is now better than OpenAI's ChatGPT. Mr Musk said Grok 3 is the "smartest AI on Earth", and the company said it performed better in benchmark tests across science, maths and coding than ChatGPT, as well as Google's Gemini, DeepSeek and Anthropic's Claude. In a livestream announcing the new model, the billionaire SpaceX and Tesla boss said Grok 3 has "more than 10 times" the compute power of Grok's previous model. The AI start-up confirmed Grok 3 will be made available to Premium+ subscribers on X - the social media platform also owned by Mr Musk - and will also be part of a new subscription called SuperGrok which will include access via the Grok mobile app and website. Mr Musk's claim about Grok being able to outperform ChatGPT - which have not been independently verified - further intensifies his rivalry with OpenAI and its chief executive, Sam Altman. Mr Musk has previously filed two lawsuits against OpenAI, claiming it has strayed from its founding principles of being a non-profit, and last week launched a bid to buy the company for 97.4 billion US dollars (£77.3 billion) - a bid that was rejected by Mr Altman and the OpenAI board. He helped co-found OpenAI but left in 2018 and has become increasingly critical of the firm and its leadership. Mr Altman dismissed the bid as a tactic to slow down a rival in the AI space, which remains centred on major firms based in the US, although DeepSeek's sudden appearance last month as a cheaper, China-based alternative to the US giants has shaken that perception. However, this has not yet appeared to slow the appetite to invest in AI firms - there are reports that OpenAI is in talks to raise around 40 billion dollars (£31.7 billion) in a new funding round which would push its valuation up to around 300 billion dollars (£238 billion). Meanwhile, it has been reported that xAI is in talks to raise around 10 billion dollars (£8 billion), which would value the firm at around 75 billion dollars (£59.5 billion).

Grok 3: Musk unveils 'smartest AI' to beat Google, OpenAI, DeepSeek

Elon Musk's artificial intelligence company, xAI, has officially released the Grok 3 series of models. The lineup includes a GPT-4o-like pre-trained model, two enhanced reasoning models, and a newly developed AI agent called Deep Search. Grok 3 is designed to operate on xAI's Colossus supercomputer, which is reported to contain over 100,000 Nvidia GPU hours dedicated to AI model training. This high-performance computing infrastructure was constructed in just over eight months to provide an advanced foundation for xAI's latest AI breakthroughs.

xAI launches Grok 3 AI, claiming it is capable of 'human reasoning'

xAI has launched its Grok 3 models during a livestream with Elon Musk, who said they were "an order of magnitude more capable than Grok 2." The Grok 3 mini model can answer questions quickly, but it's not as accurate as the other models in the family. Meanwhile, the Grok 3 Reasoning and Grok 3 mini Reasoning models are capable of mimicking human-like reasoning when it comes to analyzing information the user needs. Other examples of AI models capable of reasoning tasks are DeepSeek's R1 and OpenAI's o3-mini. According to TechCrunch, xAI claimed during the event that Grok 3 Reasoning performed better than the best version of o3-mini on several benchmarks. Grok 3's features will initially be available to subscribers paying for X's Premium+ tier, which now costs $40 a month in the US. (X raised the Premium+ tier's pricing from $16 to $22 in December -- now, less than two months later, it's almost twice as expensive.) They will also be available through an upcoming separate subscription option for the standalone Grok app and Grok on the web. Based on leaked information, the subscription option will be called SuperGrok and will cost $30 a month. With the Grok 3 models enabled, users will be able to ask the chatbot to "Think" if they want to tap its reasoning capabilities for mathematics, science and programming questions. For even more complex queries, they can use the "Big Brain" function that requires additional computing. The models' reasoning capabilities power a new Grok feature called DeepSearch, which xAI describes as the "next generation search engine." DeepSearch will scan the internet and X, formerly Twitter, to conjure a brief summary for research inquiries. In addition to launching the Grok 3 models, xAI also revealed during the event that the Grok app will get a "voice mode" within a week, giving it synthesized voices to converse with users. Grok 2, the company's older models, will be open sourced in the coming months.

Grok 3 is Kind Of Scary Smart & It's Releasing Soon: Elon Musk

Elon Musk said Grok 3 is in the final stages of polishing and could be released in a week or two. The AI industry is in a frenzy. While Apple is partnering with Alibaba for AI features, and Hugging Face is releasing a new small language model, Elon Musk's Grok AI model may have lost its charm - but not for long. At the World Governments Summit in Dubai, Musk dropped a scoop on xAI's next-generation chatbot, Grok 3, the successor to Grok 2. Not only did he reveal the expected release timeline, he also hinted at the potential performance of the chatbot. To start with, Musk said, "Grok 3 is outperforming anything that's been released that we're aware of...At times, I think Grok 3 is kind of scary smart." He revealed that the chatbot offers solutions that are not obvious. In the discussions, he shared that Grok 3 was trained with the most compute power and synthetic data. The AI model goes back and forth through the data and tries to achieve logical consistency. If it gets data that is wrong, the chatbot will reflect upon that and remove the data in question. "Its base reasoning is very good; in fact, even without fine-tuning Grok 3, the base model is better than Grok 2," he highlighted. Addressing the anticipation, he mentioned that Grok 3 is in the final stages of polishing and that it might be released in another week or two. Even with the expected timeline, he emphasised that he does not want to be hasty in the release because the final polish is necessary for a great user experience.

Elon Musk's xAI Launches Grok 3 Model It Claims Outperformed Rivals in Blind Tests

Early access to Grok 3 is available to X Premium+ subscribers starting Tuesday, as well as through a standalone website and app. Elon Musk's xAI launched Grok 3, the latest version of its AI model, which the company said outperformed rivals in blind tests. The latest model is "an order of magnitude more capable" than its predecessor, Musk said during a demonstration streamed on the X social media platform late Monday. Early access to Grok 3 is available to X Premium+ subscribers in the U.S. starting Tuesday, the company said, as well as through a standalone website and app. Voice interaction with Grok 3 is expected at some point in the future, Musk said. The Grok 3 team said an early version of its model scored better than competing models, including iterations of DeepSeek and OpenAI models, in a series of blind tests. Musk's xAI also unveiled Deep Search, which it called a "next generation search engine" that Musk said can cut down on the amount of time users spend searching for information online. The launch comes after reports last week that xAI is meeting with investors about a potential $10 billion fundraising round that would value the company at $75 billion. The company is also reportedly nearing a deal with Dell Technologies (DELL) to buy more than $5 billion of servers powered by Nvidia (NVDA) GB200 chips. Musk, who owns xAI and the X platform, recently submitted a $97.4 billion offer along with an investment group to buy the nonprofit that controls xAI rival OpenAI. The offer was rejected, first on X by CEO Sam Altman, and then by the company's board of directors last week. Musk is also leading President Trump's new Department of Government Efficiency, on top of his CEO positions at Tesla (TSLA) and SpaceX.

Grok-3 Launch Ignites AI Arms Race as OpenAI Readies Counterpunch With GPT-4.5 - Decrypt

Is it FUD or just another escalation of the AI Race? Claiming that it's "scary smart," Elon Musk plans to unveil Grok-3 tonight putting xAI's massive computing infrastructure to the test against industry leaders like OpenAI and Anthropic. But his post on X only seemed to provoke Musk's blood rival, OpenAI CEO Sam Altman, who responded today by claiming his upcoming model, GPT4.5, is tantalizingly close to AGI. But how close is it to being released? Altman had recently challenged Musk to put up or shut up: "Just compete by building a better product," he told Bloomberg. And tonight Musk seems eager to prove he can do exactly that. Musk has hyped his model and said it's "outperforming anything that has been released (to date)." Of course, experts and AI enthusiasts took this literally, especially considering xAI's models have been pretty underwhelming. If it turns out to be true, Grok-3 would put a lot of pressure on the AI industry. That is, until OpenAI 4.5 drops. "We're about to figure out what a cracked team can ship with a significantly bigger cluster than GPT-4 was trained on," wrote Mark Teneholtz, head of AI at the forecasting platform Predelo. Tenenoltz was referencing the Colossus Supercluster with 100,000 Nvidia HGX H100s GPUs that xAI used to train its upcoming models. Initially, Altman said GPT-4.5 could drop within weeks, though given his tweet early today, perhaps OpenAI is moving up the timeline to steal Musk's thunder. Already embarrassed by China's DeepSeek and with billions of dollars in game, OpenAI can't afford to sit idle. When a Twitter user urged Sam Altman to beat Grok-3 with a surprise launch, Altman replied that "wouldn't be very nice." GPT-4.5, codenamed Orion, represents OpenAI's last "non-chain-of-thought" model before the company unifies its AI capabilities in GPT-5. It's been trained using synthetic data from the o1 reasoning model, which could explain the AGI-like qualities Altman reported. Grok-3, being so powerful at reasoning, seems to follow the same path. But not everyone's buying the hype. Benjamin De Kraker, an xAI employee quit last week after the company threatened to fire him for ranking Grok-3 below ChatGPT-o1 and ChatGPT-o3 in coding performance. The ranking, which he stressed was his personal opinion, sparked immediate backlash from xAI leadership -- despite Elon Musk saying it was weird. A prominent leaker known as "@Iruletheworld," who got famous for actively leaking and posting details about Strawberry and other AI models, claimed Grok-3 would "blow your goddamn mind." The account later doubled down, declaring it represents artificial general intelligence (AGI). "It doesn't guess. It doesn't fumble logic. It doesn't need training wheels. it thinks. it reasons. it plans. when you talk to it, you feel it -- the weight of real intelligence staring back at you," he posted on X. "This is what happens when you push scale to the limit. This is what happens when you stop being cautious and just go for it." And xAI employee Frederik Meringdal went even further, suggesting to the leaker that Grok-3 would be good enough to make him change his bio, assuring Grok-3 is beyond AGI, reaching Artificial Super Intelligence. The rivalry between Musk and Altman has deep roots. After co-founding OpenAI together in 2015, Musk left in 2018 when the company rejected his proposal to merge with Tesla. Earlier this year, Musk's investor group tried to acquire OpenAI with a $97.4 billion bid, which Altman rejected. Musk said he'd withdraw if OpenAI returned to its non-profit roots. OpenAI maintains that Musk's departure stemmed from his unsuccessful attempt to take control as CEO. The dispute has since spawned legal action, with Musk suing OpenAI over its Microsoft partnership, claiming it betrayed the original non-profit mission. Perhaps if one of their companies reaches AGI or Super Intelligence, it can settle this mess.

Elon Musk's startup rolls out new Grok-3 chatbot as AI competition intensifies

Billionaire CEO claims bot is 'maximally truth-seeking' as he looks to rival DeepSeek, OpenAI and Google Gemini Elon Musk's artificial intelligence startup xAI has introduced Grok-3, the latest iteration of its chatbot that integrates with X, formerly Twitter. Grok-3 debut comes at a critical moment in the AI arms race as Musk looks to compete with the Chinese AI firm DeepSeek, Microsoft-backed OpenAI and Google. Musk's bot has seen less widespread adoption than DeepSeek's namesake chatbot, which wowed the world weeks ago and caused panic in stock markets, as well as OpenAI's ChatGPT and Google's Gemini. Grok-3 is being rolled out immediately to Premium+ subscribers of X, the social media platform owned by Musk. xAI is also launching a new subscription tier, SuperGrok, for users accessing the chatbot via its mobile app and Grok.com website. The chatbot can generate texts and images without many of the common guardrails against sexually suggestive imagery, vulgarity or the reproduction of well-known people's likenesses. X users have deployed the chatbot to mock political figures, including Musk himself, create deepfakes of celebrities and manipulate copyrighted material. "Grok-3 across the board is in a league of its own," Musk said during a livestream alongside three xAI engineers late on Monday. He added the new model outperforms its predecessor, Grok-2, boasting of "more than 10 times" the computing power of the previous version and passing AI industry benchmark tests with flying colors. He called the bot "maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct". The billionaire CEO regularly spreads falsehoods to his 200m followers on X. Musk said the bot leverages "Big Brain" mode for more complex research tasks than a normal chatbot could conduct. The latest release introduces a smart search engine, called DeepSearch, which xAI describes as a reasoning-based chatbot capable of articulating its thought process when responding to user queries. The tool, demonstrated during the livestream, offers functions for research, brainstorming and data analysis. Two weeks ago, OpenAI released a large language model that can similarly comb through the open internet and conduct more human-esque research for paying subscribers to ChatGPT. Google has also added research functions to Gemini. "The introduction of Grok-3 puts xAI back in the race for leadership in open-source LLMs [large language models]. It outperforms the current state-of-the-art models on some benchmarks, which makes xAI relevant again" said Gil Luria, managing director at DA Davidson. As competition in AI intensifies, xAI is ramping up its data center capacity to train more advanced models by raising billions of dollars. Musk touts its supercomputer cluster in Memphis, Tennessee, called "Colossus", as the largest in the world. However, Luria said improvements over the Grok-2 model appear to be too small to justify the enormous resources used to train it. Last week, a consortium of investors led by Musk offered $97.4 bn to acquire OpenAI's non-profit assets, an offer the ChatGPT-maker rejected. Aside from xAI, Musk himself has been engaged in a rapid reformation of the US federal government under Donald Trump. Asked on the day of Grok-3's release to describe his efforts in Washington, the chatbot responded: "Musk's involvement in the federal government has thus far been characterized by rapid, sweeping changes aimed at efficiency, but it has also sparked controversy over the methods, legality, and ethics of his approach. His actions are part of a broader Trump administration agenda to reduce the size and scope of government but have been uniquely aggressive and controversial due to Musk's personal involvement."

Musk launches Grok 3 AI chatbot. Here's what to know.

Elon Musk has debuted a new version of his artificial intelligence chatbot, called Grok 3, days after rival OpenAI rejected the billionaire's bid for the company, which he cofounded with Sam Altman and others in 2015. Musk's startup, xAI, claimed in a livestream announcement on Musk-owned social media platform X that Grok 3 is superior to competitors including OpenAI's GPT-4o, Alphabet's Google Gemini, DeepSeek's V3 model and Anthropic's Claude. With its latest AI chatbot, xAi's mission according to Musk, is to "understand the universe," including by answering questions such as, "Where are the aliens?" "How does the universe end?" and "How did it start?" Grok 3 surpasses the computer power of its previous iteration, Grok 2 by more than 10 times, xAI engineers said in an hour-long, livestreamed presentation on X. The company's claims about the AI's computing power and capabilities across categories such as mathematical reasoning, science and coding have not been independently verified. During the X livestream, xAI demoed Grok 3, asking the chatbot to plot a launch from earth to planet Mars, and back to earth at a future launch date. Musk and his engineers also asked it to create a game that is a hybrid between Tetris and Bejeweled, and to "make it insanely great." The company noted that Grok 3 is learning every day, and that users can expect to see improvements in the AI every 24 hours. Andrej Karpathy, former director of AI at Tesla and a member of OpenAI's founding team, gave Grok 3 a "quick vibe check," posting a review of the tool on X. He praised its "state-of-the-art thinking model," which he said is as good as OpenAI's and superior to DeepSeek's, Gemini's and Claude's. He critiqued Grok 3 "sense of humor" though, and called it "overly sensitive" to ethical issues. Premium+ X subscribers, who pay $22 a month for the service, can currently access the tool. By contrast, OpenAI's GPT-4o costs $200 a month. Musk launched xAI in 2023 to compete with OpenAI, which he left in 2018 and has since criticized over its plans to restructure as a for-profit company. OpenAI last week unanimously rejected a $97.4 million takeover bid from Musk for the AI firm. "OpenAI is not for sale, and the board has unanimously rejected Mr. Musk's latest attempt to disrupt his competition," OpenAI chairman Bret Taylor said on X. "Any potential reorganization of OpenAI will strengthen our nonprofit and its mission to ensure AGI [artificial general intelligence] benefits all of humanity."

Elon Musk's xAI To Unveil Grok 3 On Monday As AI Race With ChatGPT-Parent OpenAI And China's DeepSeek Heats Up

Enter your email to get Benzinga's ultimate morning update: The PreMarket Activity Newsletter Elon Musk's xAI plans to unveil its latest chatbot, Grok 3, on Monday. What Happened: Musk took to X, formerly Twitter, and shared the development calling Grok 3 the "Smartest AI on Earth." The launch will feature a demonstration at 8 p.m. Pacific time. The announcement follows Musk's presentation at the World Government Summit in Dubai on Thursday, where he highlighted Grok 3's capabilities. The AI model, trained on synthetic data, is designed to reflect on its mistakes by revisiting data to ensure logical consistency, reported Bloomberg. Subscribe to the Benzinga Tech Trends newsletter to get all the latest tech developments delivered to your inbox. Why It Matters: Grok 3's launch coincides with a global push for advanced and cost-effective AI chatbots. China's DeepSeek, which shocked markets with a model rivaling OpenAI's ChatGPT, is rapidly hiring experts. OpenAI is also reportedly moving forward with developing its own AI chip. This could lessen its dependence on Nvidia Corporation. Sam Altman's company has also decided to simplify its AI lineup by discontinuing the standalone 'o3' model. Last year in November, xAI achieved a valuation of $50 billion, a feat that took OpenAI nearly nine years to accomplish. Meanwhile, OpenAI's board has rejected Musk's $97.4 billion acquisition proposal. Check out more of Benzinga's Consumer Tech coverage by following this link. Photo courtesy: Shutterstock Read Next: David Tepper, Rausing Family Boost Nvidia Holdings Ahead Of Stock Volatility Disclaimer: This content was partially produced with the help of AI tools and was reviewed and published by Benzinga editors. Market News and Data brought to you by Benzinga APIs

Grok 3 launch confirmed as 10 times more powerful than previous model

Elon Musk and the xAI team announced the Grok 3 AI model in an evening live stream on Monday. The team detailed that the new model is a magnitude more capable than Grok 2, indicating Grok 3 has 10 to 15 times more power than Grok 2. They also claim that Grok 3 is more powerful than its AI model competitors such as DeekSeek and Google Gemini. Recommended Videos The xAI team said it has been improving on the Grok 3 AI model of the last several months, noting that it will be very very funny, adding that it has only been 17 months since the launch of Grok 1. Updates are continuous

Elon Musk unveils Grok 3 and 'Deep Search' tool

The tech billionaire and "special government employee" unveiled his AI company's latest model during a demo livestreamed on Musk-owned X. Musk was joined by xAI co-founders Jimmy Ba and Yuhuai "Tony" Wu, and lead engineer Igor Babuschkin, who all went into detail about Grok 3's advanced reasoning, speed, and increased training. "We're very excited to present Grok 3, which is, we think, an order of magnitude more capable than Grok 2 in a very short period of time," Musk said. "Our team's been working extremely hard over the last few months to improve Grok as much as we can, so we can give all of you access to it." The new model, which follows Grok-2, is now available for all paying subscribers with a Premium+ X account. In the demo, Ba said xAI said Grok 3 had been tested under the codename "chocolate" on the LMSYS leaderboard -- a system for testing large language models (LLMs) -- and claimed even early versions of Grok 3 were outperforming rivals like Gemini-2 Pro, DeepSeek-V3, Claude 3.5 Sonnet, and GPT-4o. To demonstrate Grok 3's reasoning capabilities in physics, Babuschkin asked the AI chatbot to "generate code for an animated 3d plot of a launch from earth landing on mars and then back to earth at the next launch window." You can watch the "unscripted" result around 30 minutes into the video above. The team also had Grok 3 write code to fuse Tetris and Bejeweled into a new game that works on its own, which Musk praised as "the beginning of creativity." During the demo, Ba also announced that X was adding a new "Deep Search" tool as a "next generation search engine" powered by Grok. The product, he said, is one that "not just helps the engineers and research and scientists to do coding, but actually helps everyone answer questions that you have day to day." Ba searched "when is the next Starship launch?" using Deep Search, to which it generated the answer Feb. 24 -- notably, Wikipedia was one of the first sources it used. "It might be sooner," Musk commentated.

Musk: New Version of Grok AI Tool Launching Monday | PYMNTS.com

Elon Musk says his xAI will debut the latest version of its Grok model this week. "Grok 3 release with live demo on Monday night at 8pm PT," Musk wrote on his X social media platform Saturday (Feb. 15) evening. "Smartest AI on Earth." In another post on Sunday (Feb. 16), Musk shared some of Grok's handiwork, a play on one of the poems from "The Lord of the Rings" books, reimagined to introduce "creation of advanced large language models (LLMs) with search, agency, memory, and humor." "I've kept the tone, structure, and gravitas as close to the original as possible while aligning it with the technological theme," he added. In another post, Musk shared the results of a query in which he apparently asked Grok its opinion on the tech/business publication The Information. Grok's answer, too long to get into here, is negative, calling X the only place for "trustworthy" news. "Grok is so based," the 53-year-old Musk wrote -- using internet slang for "not caring what others think" -- adding the "tears of laughter" emoji. According to a report by Bloomberg News, Musk had teased the planned launch of Grok 3 at the World Government Summit in Dubai last week, claiming it would outdo every competing model introduced thus far. Musk added that the model was trained on synthetic data and can reflect on mistakes that it makes by reviewing data to achieve logical consistency. Last month, xAi debuted a stand-alone consumer app called Grok, offering access to the company's chatbot of the same name. The chatbot had previously only been available to users of X (formerly Twitter). The latest news comes as Musk -- in addition to his high-level White House activities -- is also engaged in a legal battle with OpenAI, the artificial intelligence (AI) startup he helped found in 2015, and Sam Altman, OpenAI's CEO. Musk has sued OpenAI to stop the company's switch to a for-profit entity (it is now controlled by a nonprofit), something he says is a betrayal of its founding ethos. Last week, the multibillionaire and a group of investors made a $97.4 billion bid for control of OpenAI. Days later, Musk said that bid would be withdrawn if the company agreed to halt its conversion to a for-profit. Altman rejected Musk's overtures, and the OpenAI board did the same days later. "OpenAI is not for sale, and the board has unanimously rejected Mr. Musk's latest attempt to disrupt his competition," board chair Bret Taylor wrote on X, speaking for the entire board. "Any potential reorganization of OpenAI will strengthen our nonprofit and its mission to ensure AGI [artificial general intelligence] benefits all of humanity."

'Smartest AI on Earth': Elon Musk's xAI unveils 'scary smart' Grok 3; All you need to know

Elon Musk's XAI unveiled its latest artificial intelligence chatbot, Grok 3, which the tech billionaire has described as the "smartest AI on Earth". Musk's AI startup xAI launched its Grok 3 chatbot today at 9.30 am. The launch event, held on February 17, 2025, showcased the capabilities of Grok 3, including advanced reasoning, text-to-video conversion, and self-correction mechanisms. At the launch event, Musk explained the meaning of name "Grok". He said the term "grok" comes from Robert Heinlein's science fiction novel "Stranger in a Strange Land." In the novel, the word "grok" is used by a character raised on Mars and means to fully and profoundly understand something. Musk emphasised that the word conveys deep understanding and empathy, which are key attributes of Grok 3.

Elon Musk to unveil Grok 3 chatbot 'smartest AI on Earth' today: Check release time and key details

Tech billionaire Elon Musk said his startup xAI will release its Grok 3 chatbot on Monday and billed it as the "smartest AI on Earth" in a fiercely competitive market. Earlier this week, the Tesla boss said that his ChatGPT challenger was in the final stages of development. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," he had said in a video call addressing the World Governments Summit in Dubai. He expressed confidence that Grok 3 will outperform existing AI models, including OpenAI's ChatGPT. ALSO READ: Elon Musk's unique gift to PM Modi during his US visit captures netizens' attention. What is it? -World's richest man Elon Musk has once again garnered global attention with the announcement of Grok 3, the latest iteration of xAI's AI chatbot. Musk described it as the "smartest AI on Earth," with a live demonstration scheduled for Monday at 8 p.m. Pacific Time (PT). This corresponds to 11 p.m. Eastern Time (ET) and 9:30 a.m. Indian Standard Time (IST, Tuesday). -At the moment, not much is known about Grok 3's capabilities, but early indications suggest the AI model will introduce advanced features, possibly text-to-video conversion and significant efficiency improvement. Musk's announcement, made via his social media platform X (formerly Twitter), has amplified speculation about Grok 3's potential. ALSO READ: Kanye West-Bianca Censori Valentine's Day plan revealed amid divorce buzz days after shocking Grammys stunt -Grok 3 was trained on synthetic data and is capable of reflecting on errors it makes by going over data in order to reach logical consistency, according to news agency AFP. -If the above enhancements are available in Grok 3, then it would position itself as a major competitor to leading AI models like OpenAI's GPT-4, Google DeepMind's Gemini, and Anthropic's Claude. But it is also pertinent to note that OpenAI continues to refine its ChatGPT models, Google is pushing its Gemini AI, and Meta is expanding its LLaMA series. ALSO READ: 'So amazing, wonderful dad': Elon Musk, world's richest man, gets big praise as he takes his kids to meet PM Modi -Grok 3's development was pushed by its Colossus supercomputer. Built in just eight months, the system is powered by 100,000 Nvidia GPU hours for training. Musk has repeatedly warned that AI poses a risk to human civilization, but he is nonetheless pushing hard for a bigger slice of investment in the sector. -Billed as a more efficient successor to Grok 2, the upcoming rendition boasts synthetic datasets, self-correction mechanisms and reinforcement learning, according to a report in Forbes. These integrations will help reduce incorrect responses as accuracy is enhanced and training times are reduced. -However, critics are not that convinced. Experts question whether Grok 3 will actually surpass the likes of GPT-4 Turbo, which has also already showcased remarkable reasoning, problem solving and multimodal capabilities. ALSO READ: Supermassive black hole, 6,00,000 times the mass of Sun, is heading towards the Milky Way. When is the collision likely to happen? -Grok 3's release comes merely a month after Chinese startup DeepSeek shocked the global AI industry with the launch of its low-cost, high-quality chatbot -- a challenge to US ambitions to lead the world in developing the technology. DeepSeek quickly overtook ChatGPT in downloads on the Apple app store. -xAI said in December it raised $6 billion in its latest funding round from investors that included US venture capitalists, chipmakers Nvidia and AMD, and investment funds from Saudi Arabia and Qatar, among others. It raised an initial $6 billion in May. The company is now one of the world's most valuable startups, though still dwarfed by OpenAI. -Elon Musk, who also acts as boss of SpaceX and Tesla, launched the AI company in July 2023 shortly after he signed an open letter calling for a pause in the development of powerful AI models. OpenAI's board chairman on Friday said it has unanimously rejected a Musk-led offer to buy the company for $97.4 billion. (With inputs from agencies)

Musk says chatbot Grok 3 will be unveiled Monday

Elon Musk said his startup xAI will release its Grok 3 chatbot on Monday and billed it as the "smartest AI on Earth" in a fiercely competitive market. The company's flagship artificial intelligence product will go live with a demonstration on Monday night at 8:00 pm Pacific time (0400 GMT), the tech billionaire wrote Saturday on his social media platform X. Grok 3 was trained on synthetic data and is capable of reflecting on errors it makes by going over data in order to reach logical consistency. "Will be honing product with the team all weekend, so offline until then," said Musk, the world's richest person and a top advisor to President Donald Trump who is tasked with slashing government spending. Musk said last week that Grok 3 was in the final stages of development and would be released to the world in a matter of weeks. xAI is seeking a competitive edge in a market teeming with products like OpenAI's ChatGPT as artificial intelligence spreads through contemporary life. Chinese startup DeepSeek shocked the global AI industry last month with the launch of its low-cost, high-quality chatbot -- a challenge to US ambitions to lead the world in developing the technology. DeepSeek quickly overtook ChatGPT in downloads on the Apple app store. Musk has repeatedly warned that AI poses a risk to human civilization, but he is nonetheless pushing hard for a bigger slice of investment in the sector. xAI said in December it raised $6 billion in its latest funding round from investors that included US venture capitalists, chipmakers Nvidia and AMD, and investment funds from Saudi Arabia and Qatar, among others. It raised an initial $6 billion in May. The company is now one of the world's most valuable startups, though still dwarfed by OpenAI. Musk, who also acts as boss of SpaceX and Tesla, launched the AI company in July 2023 shortly after he signed an open letter calling for a pause in the development of powerful AI models. OpenAI's board chairman on Friday said it has unanimously rejected a Musk-led offer to buy the company for $97.4 billion.

Grok 3: AI chatbot, ChatGPT challenger to release soon

Image credit: Nathan Laine/Bloomberg Elon Musk said on Thursday his AI chatbot, and ChatGPT challenger, Grok 3, is in the final stages of development and will be released in about a week or two. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," he said in a video call addressing the World Governments Summit in Dubai. Read: How industrial AI is leading economic hubs toward diversification, autonomy The billionaire tech mogul founded xAI as a challenger to Microsoft-backed OpenAI and Alphabet's Google. Musk also co-founded OpenAI. On Monday, a consortium of investors led by Musk said it had offered $97.4bn to buy the assets of OpenAI's nonprofit, in another salvo from the world's richest man against the artificial intelligence startup. OpenAI has said it wants to become a for-profit organization to secure the capital needed for developing the best AI models. Important: Tabby raises $160m, becomes MENA's most 'valuable' fintech Musk sued OpenAI CEO Sam Altman and others in August and has asked a U.S. district judge to block OpenAI's attempt to transition to a for-profit entity. OpenAI said this week Musk's bid clashes with his lawsuit. "I think the evidence is there in that OpenAI has gotten this far while having at least a sort of dual profit, non-profit role. What they're trying to do now is to completely delete the non-profit, and that seems really going too far." Musk, who was appointed by US President Donald Trump to oversee the so-called Department of Government Efficiency aimed at dramatically reducing the size of the federal workforce, said government spending could be reduced by $1tn or more. Must know: Dubai Duty Free introduces new way to shop "Maybe the economy could grow at 4 or 5 per cent potentially, in terms of real useful goods and services output, and government spending can be reduced by about 3 or 4 per cent of the economy, about maybe a trillion dollars or more, and the net effect of that would be no inflation from 2025 to 2026 so that would be quite remarkable," Musk said. UAE AI Minister Omar Al Olama, who was interviewing Musk at the conference, said they would partner on "Dubai Loop", an underground high-speed transport system. Turning to international affairs, Musk told the Middle East audience the United States has been "pushy" in the past and it should "mind its own business". "I think we should, in general leave other countries to their own business," he said.

Elon Musk says 'scary-smart' Grok 3 outperforms rival AI chatbots,...

Elon Musk on Thursday claimed that his latest version of generative AI, Grok 3, is "outperforming" all rival chatbots and will be released by the end of the month. The billionaire started his artificial intelligence startup xAI in 2023 and launched Grok as a direct competitor to OpenAI's ChatGPT. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," Musk said during a video call addressing the World Governments Summit in Dubai. Musk called Grok 3 "scary-smart," saying the bot has been able to come up with "not obvious solutions" that people would not anticipate. "We think it'll be better than anything else, and then maybe this might be the last time that any AI is better than Grok," he continued, a challenge to OpenAI's long dominant ChatGPT. Musk co-founded OpenAI with its chief executive, Sam Altman, in 2015, but later ties with the firm in 2018 and has since been embroiled in a legal battle with the AI rival. On Monday, a group of investors led by Musk said it had made a shocking $97.4 billion offer to buy the assets of OpenAI's nonprofit in an attempt to stop the company from transitioning to a for-profit structure. Musk, who also runs Tesla, SpaceX and the social media platform X, sued Altman and others in August, and has asked a US district judge to block OpenAI's switch to a for-profit structure. "I think the evidence is there in that OpenAI has gotten this far while having at least a sort of dual profit, non-profit role," Musk said on Thursday. "What they're trying to do now is to completely delete the non-profit, and that seems really going too far." Altman had quickly shot down the offer, and seemingly mocked the $44 billion price Musk paid for X, formerly known as Twitter, in 2022. "No thank you but we will buy Twitter for $9.74 billion if you want," Altman wrote in a post on X. During the video call, Musk also discussed his position running the so-called Department of Government Efficiency, a newly-created task force that aims to slash federal spending. On Thursday, he claimed DOGE could reduce government spending by $1 trillion or more. "Maybe the economy could grow at 4 or 5% potentially, in terms of real useful goods and services output, and government spending can be reduced by about 3 or 4% of the economy, about maybe a trillion dollars or more, and the net effect of that would be no inflation from 2025 to 2026 so that would be quite remarkable," Musk said. Speaking on international affairs, Musk told the Dubai audience that the United States has been too "pushy" in the past and should "mind its own business." "I think we should, in general, leave other countries to their own business," he said. His comments come as President Trump has continued to float his controversial Gaza redevelopment plan, which would include displacing Palestinians and turning the strip into "the Riviera of the Middle East." The United Arab Emirates artificial intelligence minister, Omar Al Olama, who was interviewing Musk during the conference, also said the UAE and Musk would partner on "Dubai Loop," an underground high-speed transit system that Musk compared to a wormhole.

Grok 3: This AI chatbot, ChatGPT challenger is to release soon

Image credit: Nathan Laine/Bloomberg Elon Musk said on Thursday his AI chatbot, and ChatGPT challenger, Grok 3, is in the final stages of development and will be released in about a week or two. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," he said in a video call addressing the World Governments Summit in Dubai. Read: How industrial AI is leading economic hubs toward diversification, autonomy The billionaire tech mogul founded xAI as a challenger to Microsoft-backed OpenAI and Alphabet's Google. Musk also co-founded OpenAI. On Monday, a consortium of investors led by Musk said it had offered $97.4bn to buy the assets of OpenAI's nonprofit, in another salvo from the world's richest man against the artificial intelligence startup. OpenAI has said it wants to become a for-profit organization to secure the capital needed for developing the best AI models. Important: Tabby raises $160m, becomes MENA's most 'valuable' fintech Musk sued OpenAI CEO Sam Altman and others in August and has asked a U.S. district judge to block OpenAI's attempt to transition to a for-profit entity. OpenAI said this week Musk's bid clashes with his lawsuit. "I think the evidence is there in that OpenAI has gotten this far while having at least a sort of dual profit, non-profit role. What they're trying to do now is to completely delete the non-profit, and that seems really going too far." Musk, who was appointed by US President Donald Trump to oversee the so-called Department of Government Efficiency aimed at dramatically reducing the size of the federal workforce, said government spending could be reduced by $1tn or more. Must know: Dubai Duty Free introduces new way to shop "Maybe the economy could grow at 4 or 5 per cent potentially, in terms of real useful goods and services output, and government spending can be reduced by about 3 or 4 per cent of the economy, about maybe a trillion dollars or more, and the net effect of that would be no inflation from 2025 to 2026 so that would be quite remarkable," Musk said. UAE AI Minister Omar Al Olama, who was interviewing Musk at the conference, said they would partner on "Dubai Loop", an underground high-speed transport system. Turning to international affairs, Musk told the Middle East audience the United States has been "pushy" in the past and it should "mind its own business". "I think we should, in general leave other countries to their own business," he said.

Musk says Grok 3, rival to OpenAI's ChatGPT, to launch in weeks

STORY: :: Elon Musk says human intelligence will be dwarfed by machines as he talks AI chatbots in Dubai "We're really in the final stages of polishing Grok 3. Probably it gets released in about a week or two, so pretty soon." :: February 13, 2025 "We think it'll be better than anything else. And then maybe this might be the last time that any AI is better than Grok." "OpenAI has gotten this far while having at least a sort of dual profit, non-profit role. What they're trying to do now is completely delete the non-profit and that seems really going too far." "But human intelligence, I think, will be dwarfed by machine intelligence. I'm not sure how to feel about that, except that it is to be inevitable that at some point, human intelligence will be a very small fraction of total intelligence. Digital intelligence will be more than 99% of all intelligence in the future." Speaking at a video call addressing the World Governments Summit, Musk said Grok 3 - a challenger to ChatGPT - was outperforming anything else that has been released "that we're aware of." The billionaire tech mogul founded xAI as a challenger to Microsoft-backed OpenAI and Alphabet's Google. Musk also co-founded OpenAI. On Monday, a consortium of investors led by Musk said it had offered $97.4 billion to buy the assets of OpenAI's nonprofit. UAE AI Minister Omar Al Olama, who was interviewing Musk at the conference, said they would partner on "Dubai Loop," an underground high-speed transport system that Musk likened to a wormhole. Al Olama did not give details. Turning to international affairs, Musk told the Middle East audience the United States has been "pushy" in the past and it should "mind its own business."

Musk launches 'scary smart' AI chatbot

SAN FRANCISCO (AFP) - Elon Musk's artificial intelligence company unveiled on Monday the latest version of its chatbot, Grok 3, which the billionaire hopes will find traction in a highly competitive sector contested by the likes of ChatGPT and China's DeepSeek. The launch comes as the world's richest man is deploying the enormous powers granted him by US President Donald Trump to restructure and dismantle federal agencies. The unprecedented cost-cutting drive has raised conflict-of-interest questions, given that many of those agencies have regulatory oversight on elements of Musk's sprawling business empire. "Grok is to understand the universe," Musk said at the start of the Grok 3 launch presentation. "We're driven by curiosity about the nature of the universe -- that's also what causes us to be a maximally truth-seeking AI, even if that truth is sometimes at odds with what is politically correct." Musk has promoted Grok 3 as "scary smart," with 10 times the computational resources of its predecessor that was released in August last year. The flagship product of his xAI company was trained on synthetic data and employs self-correction mechanisms that avoid errors -- known as "hallucinations" -- that plague some AI chatbots and lead them to process false or misleading data as fact. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," Musk said in a video call last week with the World Governments Summit in Dubai. Grok 3 will be made available first to Premium+ paid subscribers of X -- formerly Twitter, which Musk acquired in 2022 -- before rolling out to other users. The upgraded chatbot enters a crowded field with countries racing to introduce more sophisticated -- and cost-effective -- AI products. Chinese startup DeepSeek shocked the global AI industry last month with the launch of its low-cost, high-quality R1 chatbot -- a direct challenge to US ambitions to lead the world in developing the technology. Grok 3 is also going up against OpenAI's chatbot, ChatGPT - pitting Musk against collaborator-turned-arch rival Sam Altman. Musk and Altman were among the 11-person team that founded OpenAI in 2015. Created as a counterweight to Google's dominance in artificial intelligence, the project got its initial funding from Musk, who invested USD45 million to get it started. Musk left three years later, and then in 2022 OpenAI's release of ChatGPT created a global technology sensation -- one that did not feature Musk at its center and which made Altman a star. Their relationship has become increasingly toxic and litigious ever since, with Open AI's board last week rejecting a Musk-led offer to buy out the company for close to $100 billion. Trump has put technology front and center of his new administration. Tech billionaires featured prominently at his inauguration and he has announced a number of major AI infrastructure initiatives from the White House. Musk has become a key figure in the administration, as one of Trump's closest advisers and the head of the newly created Department of Government Efficiency (DOGE), which has begun a radical overhaul of the US government bureaucracy. Critics warn that Musk's proximity to the president poses a major conflict of interest as he guides Trump on laws and regulations around artificial intelligence -- just one sector in which he has a substantial commercial stake. According to Bloomberg, xAI has been canvassing potential investors for a roughly USD10 billion funding round that would value the company at about USD75 billion. Musk, who also acts as boss of SpaceX and Tesla, launched the xAI company in July 2023, shortly after he signed an open letter calling for a pause in the development of powerful AI models.

Musk Says xAI's 'Scary Smart' Next Model Chatbot Coming in Weeks

Elon Musk praised his upcoming Grok 3 chatbot as an AI model outperforming everything else that's been released thus far, and said the world would get to see it in a matter of weeks. "At times I think Grok 3 is scary smart," Musk, the billionaire entrepreneur and now close adviser to US President Donald Trump, said via video conference at the World Government Summit in Dubai on Thursday.

Elon Musk says Grok 3 in final stages, outperforming all chatbots

* Musk says OpenAI 'going too far' in deleting non-profit segment * Tells Middle East audience that US should mind its own business * US government spending could be cut by $1 trillion and reduce inflation to zero, Musk says DUBAI, Feb 13 (Reuters) - Elon Musk said on Thursday his AI chatbot, and ChatGPT challenger, Grok 3, is in the final stages of development and will be released in about a week or two. "Grok 3 has very powerful reasoning capabilities, so in the tests that we've done thus far, Grok 3 is outperforming anything that's been released, that we're aware of, so that's a good sign," he said in a video call addressing the World Governments Summit in Dubai. The billionaire tech mogul founded xAI as a challenger to Microsoft-backed OpenAI and Alphabet's Google. Musk also co-founded OpenAI. On Monday, a consortium of investors led by Musk said it had offered $97.4 billion to buy the assets of OpenAI's nonprofit, in another salvo from the world's richest man against the artificial intelligence startup. OpenAI has said it wants to become a for-profit organization to secure the capital needed for developing the best AI models. Musk sued OpenAI CEO Sam Altman and others in August and has asked a U.S. district judge to block OpenAI's attempt to transition to a for-profit entity. OpenAI said this week Musk's bid clashes with his lawsuit. "I think the evidence is there in that OpenAI has gotten this far while having at least a sort of dual profit, non-profit role. What they're trying to do now is to completely delete the non-profit, and that seems really going too far." Musk, who was appointed by U.S. President Donald Trump to oversee the so-called Department of Government Efficiency aimed at dramatically reducing the size of the federal workforce, said government spending could be reduced by $1 trillion or more. "Maybe the economy could grow at 4 or 5% potentially, in terms of real useful goods and services output, and government spending can be reduced by about 3 or 4% of the economy, about maybe a trillion dollars or more, and the net effect of that would be no inflation from 2025 to 2026 so that would be quite remarkable," Musk said. UAE AI Minister Omar Al Olama, who was interviewing Musk at the conference, said they would partner on "Dubai Loop", an underground high-speed transport system that Musk likened to a wormhole. Al Olama did not give details. Turning to international affairs, Musk told the Middle East audience the United States has been "pushy" in the past and it should "mind its own business". "I think we should, in general leave other countries to their own business," he said. Trump has enraged the Arab world by saying the U.S. would take over the Gaza strip, resettle its Palestinian inhabitants and turn it into the "Riviera of the Middle East". (Reporting by Yousef Saba; writing by Maha El Dahan; Editing by Himani Sarkar and Sonali Paul)