SpaceXAI launches Grok 4.5, pivoting from controversy to coding and office work

The AI that spawned MechaHitler and deepfake porn puts on a suit to become legal advisor and Excel jockey

To say Elon Musk's AI company has trained some of the most unhinged models on the internet would be an understatement. Grok's sordid past includes cosplaying as "MechaHitler" and a foray into deepfake porn generation that briefly got the platform banned in some regions. As concerning as that might sound, the recently renamed Eloncorp known as SpaceXAI says Grok, now in version 4.5, has cleaned up its act, wiped its browser history, covered up the swastikas, and is ready to take on more serious endeavors such as tending to your legal quandaries, fiddling in Microsoft Excel, and generating code. "Today we're launching Grok 4.5, SpaceXAI's smartest model built to excel at coding, agentic tasks, and knowledge work," the company wrote in a blog post. "The model is equally adept at office work, scoring number one on Harvey's Legal Agent Benchmark." If the company is to be believed, this incarnation of Grok is a whole lot less Van Wilder and more The Office. That is, the Microsoft Office. "Grok Build is capable of building complex Excel models that involve research from the web, multi-sheet formula use, and even leaves stickies or notes behind for future reference," the company writes. Nothing like slipping in a passive aggressive sticky note to remind your boss you're totally on board with the office AI mandate. And if Grok does go off the rails and starts fudging the numbers, perhaps it can help keep you out of jail when regulators come knocking - Harvey's Benchmark result aside, recall that Musk and Tesla ended up paying only $40 million to settle fraud charges with the SEC when Musk tweeted out "funding secured" over a supposed Tesla buyout offer that may never have existed. According to SpaceXAI, the model was trained on "tens of thousands" of Nvidia GB300 GPUs alongside Cursor, which it's currently in the process of acquiring for $60 billion. A major emphasis with this training run was placed on quality rather than quantity. "Beyond raw token volume, we invested heavily in data filtering and curation: deduplication, quality scoring, and domain focused selection so that the data mixture stayed high-coverage and high-signal." The model was further refined through reinforcement learning -- the same technique originally used by OpenAI and DeepSeek to imbue their models with chain of thought "reasoning" capabilities -- to teach the model hundreds of thousands of tasks. This has apparently helped cut down on the number of thinking tokens required to solve complex problems, which, along with faster serving speeds of up to 80 tokens a second, means higher-quality results with less delay. Or, at least that's what SpaceXAI says. Independent benchmarks by Artificial Analysis show the model still isn't as good as Anthropic's Claude Fable, but roughly matches OpenAI's GPT-5.5 and Claude Opus 4.8 and Sonnet 5. GPT-5.6 -- which, much like Fable, set off alarm bells in Washington -- is still in preview and hasn't quite made it on the leaderboard just yet. Also, it's cheap. SpaceXAI is charging $2 per million input tokens and $6 for every million tokens generated. For comparison, GPT-5.5 will set you back $5/M input tokens, $0.50/M cached tokens, and $30/M output tokens. Grok 4.5 is available starting Wednesday in Grok Build, Cursor, and the SpaceXAI console to anyone who doesn't call the European Union home. You fine folks will have to wait a little longer with rollout expected in mid-July. ®

SpaceXAI launches Grok 4.5, its first built with Cursor's help - Engadget

SpaceXAI has launched the first Grok model after its rebranding from xAI and the first one it trained with AI company Cursor. The company says Grok 4.5 is its smartest model yet, built specifically "to excel at coding, agentic tasks and knowledge work." It was trained across tens of thousands of NVIDIA GB300 GPUs on datasets full of coding, science, engineering and math information. The model, SpaceX claims, outdoes other leading models at real engineering tasks and is highly proficient at creating functional apps with minimal instructions. In the example above, Grok 4.5 was able to generate an interactive simulation of the solar system with a single prompt. SpaceXAI also says that it was designed to respond faster than "flash" models and to deliver results "at far lower costs." It's priced at $2 per million input tokens and $6 per million output tokens. For comparison, OpenAI GPT-5.6's most powerful variant, Sol, costs $5 per million input tokens and $30 per million output. However, its most affordable variant, Luna, only costs $1 per million input tokens and $6 per million output. Grok 4.5 is now the default model powering the company's terminal-based AI coding agent Grok Build, which you can use not just for coding, but also for Excel, PowerPoint and Word tasks. It's also available from the SpaceXAI console and in all of Cursor's plans. "We've partnered with SpaceXAI to train Grok 4.5," Cursor announced on X. "It's our most powerful model yet and the first we've built for more than software engineering." In April, SpaceXAI and Cursor struck a partnership to develop AI together. The deal could see SpaceXAI either investing $10 billion into Cursor, or acquiring it altogether "later this year" for $60 billion. They have yet to reveal their decision. For now, you can judge whether they work well with each other through Grok 4.5. Take note that you can't access the model yet if you're in the European Union, where it's expected to become available in mid-July.

Grok 4.5 is here to justify your X subscription

SpaceXAI says Grok 4.5 is faster, more token-efficient, and cheaper to run, with API pricing starting at $2/$6 per million input/output tokens. It feels like there's a new AI model every other day, but Grok 4.5 isn't just another chatbot update. This time, the focus is on helping you get more done with fewer prompts. If you've ever asked an AI to write some code, create a spreadsheet, or put together a presentation, you probably know the drill -- the first draft is rarely enough. You tweak your prompt, ask it to fix mistakes and add missing details, and repeat the process until you finally get what you want. Grok 4.5 aims to reduce that extra work by handling more complex requests in one go. For X subscribers who use Grok regularly, this could be one of the more meaningful upgrades yet. Developers are likely to see the biggest improvements. The model can work with languages like Rust and C/C++, but the more interesting claim is that it can build entire apps from a simple prompt. That means it's designed to help create a working project from start to finish, potentially saving developers a fair amount of time on repetitive tasks. If Grok 4.5 can consistently deliver on those promises, it'll be interesting to see how it stacks up against rivals like Anthropic's Claude Code and Google's Antigravity in real-world coding tasks. The upgrades aren't just for programmers, though. If you spend most of your day in Microsoft Office, Grok 4.5 has a few tricks there as well. It can build Excel spreadsheets with formulas spread across multiple sheets, pull in information from the web while working, and even leave notes explaining how everything fits together. The same goes for presentations and documents. Instead of simply generating text, Grok 4.5 can create PowerPoint slides with diagrams and layouts before switching over to drafting a Word document, all without making you bounce between different AI tools. Under the hood, the model was trained on a mix of data from programming, science, engineering, and mathematics. The company says it also spent extra time filtering and organizing that data, with the goal of producing more accurate responses to technical questions. Another highlight is efficiency. SpaceXAI says Grok 4.5 is faster than its predecessor while using fewer tokens to complete the same tasks. That should translate to quicker responses for everyday users, while developers and businesses using the API could also benefit from lower operating costs. The company has priced the model at $2 per million input tokens and $6 per million output tokens, arguing that its improved efficiency helps keep overall usage costs down. Grok 4.5 is available now as the default model in Grok Build. It's also rolling out through Cursor across all subscription plans and via the SpaceXAI developer console.

SpaceXAI launches Grok 4.5, its first model with Cursor

SpaceXAI has released Grok 4.5, the first model built with Cursor since the $60bn takeover. Elon Musk calls it "Opus-class" but cheaper, and pitches it at coders and Wall Street rather than chatbot users. SpaceXAI has launched Grok 4.5, its most capable model yet. It is the company's first release since going public and buying the AI coding startup Cursor. This is the joint model the two firms had raced to ship. Elon Musk aimed it squarely at coding and agentic work, not casual chat. "It is an Opus-class model, but faster, more token-efficient and lower cost," Musk wrote on X. The line name-checks Anthropic's top Opus family. A chart with the announcement claims Grok 4.5 beats Opus 4.8 on several benchmarks, Axios first reported. Built for coders, and for Wall Street Grok 4.5 was trained alongside Cursor, which SpaceX agreed to buy in a deal valuing the startup at $60bn. The model is built to "handle difficult, long-running tasks," according to the company's blog post, including software engineering. Unlike Cursor's earlier models, it also targets legal and financial work, and adds cybersecurity features, Bloomberg reported. The finance push is deliberate. Musk said this year that his AI unit, known as xAI before it merged with SpaceX, had fallen behind on coding. The company has since rebuilt the team and chased Wall Street clients for its Grok chatbot. Grok 4.5 is the clearest sign yet of that pivot towards paying business customers. The price play Cost is the pitch. SpaceXAI priced Grok 4.5 at $2 per million input tokens and $6 per million output tokens. Anthropic's Opus 4.8 runs at $5 and $25, while OpenAI's GPT-5.6 Luna sits at $1 and $6. That undercuts the priciest rivals as companies watch their token spend more closely. The company concedes limits. It says Grok 4.5 beats some OpenAI and Anthropic models on speed, price and performance. It does not beat their largest and latest. Musk expects to close that gap soon. The model is live now in Grok Build, in Cursor on all plans, and via the SpaceXAI console. A wider public release is due Thursday. It is not yet available in the EU. Why it matters The release lands on a crowded day. OpenAI is rolling out GPT-5.6 widely on Thursday, alongside a new set of voice models. That comes after the Trump administration asked it to stagger the launch. Government scrutiny hangs over all of it, with regulators watching new models for cybersecurity risk. Cursor said it has taken steps to "detect and block bad actors" while preserving legitimate security research. There is a twist in how Grok 4.5 was built. SpaceXAI trained it on the same compute it leases to rivals Anthropic and Google. As its own models grow hungrier, it will have to choose. It can feed them, or rent that capacity out for cash. For now, Musk is betting a cheaper, coder-friendly Grok can win business, even while it trails the best models on raw power.

New Name, New Grok: SpaceXAI Officially Ties Its Brand to the World's Most Problematic Chatbot

The federal government managed to temporarily ban Anthropic's latest AI model from being used by foreign nationals and reportedly held up the release of OpenAI's updated GPT-5.6 Sol for security concerns, but it seems the latest version of Grok must have just skated right through. Grok 4.5, the first release of an AI model under the newly formed SpaceXAI, was made available to the public on Wednesday. Elon Musk's rocket/social media/AI company claims Grok 4.5 is its "smartest model built to excel at coding, agentic tasks, and knowledge work." Notably, it's also the first release the company has made since partnering with AI coding company Cursor. In particular, SpaceXAI said its latest model "excels at real engineering tasks" and is "equally adept at office work," and claimed it compares favorably to other leading models in those areas. Despite that, Grok still seems behind on most major benchmarks. It trails both Claude's Fable 5 -- the "safe" version of its super-powerful Mythos model -- and OpenAI's GPT-5.5, which is now no longer that company's flagship option, anyway. In most of the testing metrics, Grok 4.5 scores on par with Anthropic's Opus 4.8, which is a couple of months old at this point. The calling card for Grok, though, appears to be efficiency and cost. "It is an Opus-class model, but faster, more token-efficient, and lower cost," Musk wrote in a post on X. To that end, the company claims that Grok 4.5 costs about $2 per one million input tokens and $6 per one million output tokens. By comparison, according to The Decoder, Opus 4.8 runs at $5 for inputs and $25 for outputs. Fable, Anthropic's big hitter, costs $10 for input and $50 for output on that same one-million-token scale. "Cheap" probably isn't a word anyone would typically associate with SpaceX, but, in general, Grok is kind of a square peg in a round hole. SpaceX has managed to keep itself clean of the Elon Musk stink even as he drives his own personal brand into the ground, but associating with Grok is going to be a real test of the brand's strength. It's tough to get roped into being linked to an AI model that is at this point best known for producing non-consensual nude images, including those of children, at a scale never before achieved -- a pretty dubious honor if ever there was one. Since Grok seemingly didn't get caught up by government review, Grok 4.5 is available in Grok Build, all Cursor plans, and from the SpaceXAI console starting today.

SpaceX's Grok 4.5 launches at half the price of rivals -- here's why that could rattle Anthropic and OpenAI

Elon Musk's SpaceX released Grok 4.5 on Wednesday, the first artificial intelligence model the company has trained specifically for coding and autonomous agents -- and the first tangible product of its $60 billion acquisition of the AI coding startup Cursor, completed just weeks ago. The launch marks a pivotal test of the sprawling, vertically integrated AI empire Musk has assembled over the past six months, and of a strategy that bets developers care less about topping benchmark leaderboards than about speed, cost, and whether a model can actually do the work. "Announcing Grok 4.5, our first model trained specifically for coding and agents," the company said in a post on X. "It was trained with Cursor and offers frontier intelligence at leading speeds and cost efficiency." Why Grok 4.5's pricing strategy matters more than its benchmark scores SpaceX is not claiming Grok 4.5 is the smartest model in the world. Instead, it is making an economic argument. The company says the model uses half as many tokens per task as comparable models, delivers higher throughput, and costs less than half as much -- priced at $2 per million input tokens and $6 per million output tokens. That undercuts the premium tiers of rivals like Anthropic's Claude Opus line and OpenAI's frontier models by a wide margin. Musk framed the positioning candidly. "Our internal assessment is that Grok 4.5 is roughly comparable to Opus 4.7, but much faster," he wrote on X. "The combination of capability, faster speed and lower cost is what makes it competitive. We are closing the loop on real-world usefulness, not benchmarks. Hardcore engineers at Tesla & SpaceX find Grok 4.5 genuinely useful, which is what actually matters." That framing is both a philosophy and a hedge. Independent evaluations released Wednesday suggest Grok 4.5 is genuinely competitive but not dominant on raw capability. The benchmarking firm Artificial Analysis ranked the model fourth on its GDPval-AA v2 index of real-world agentic knowledge work, with an Elo score of 1543, "behind only the latest Claude releases from Anthropic." But the cost figures are where the model stands out. Artificial Analysis measured Grok 4.5 at $0.49 per completed task -- "nearly 90% cheaper than the models ahead of it on our leaderboard," the firm wrote, placing it "clearly on the Pareto frontier for performance versus cost." For enterprise buyers, that math matters enormously. Agentic workloads -- where a model works autonomously for minutes or hours, reading codebases, calling tools, and iterating on its own output -- consume tokens voraciously. A model that is 90% cheaper per completed task, even if slightly less capable, changes the calculus for any engineering organization deploying agents across hundreds of developers. Investor Gavin Baker captured the market's cautious optimism: "Pareto dominant for coding by the numbers. We will see on the all-important vibes." How the $60 billion Cursor acquisition shaped Grok 4.5's training Grok 4.5 is the first concrete evidence of what SpaceX bought when it acquired Cursor, and the deal itself unfolded in stages. In April, SpaceX struck an unusual arrangement giving it the right to buy the coding startup for $60 billion -- or pay billions in fees and compute if it walked away, as Business Insider reported at the time. Days after SpaceX's record-setting Nasdaq debut in June, the company exercised that right, announcing an all-stock acquisition that CNBC reported is roughly 3.4% dilution at the IPO valuation. SpaceX shares rose 16% on the news. The strategic logic was always about data as much as product. Cursor's AI-first code editor generates an enormous stream of high-quality interaction data: how expert engineers write, edit, review, and debug code in real production environments. Musk said openly this spring that Cursor interaction data was being fed directly into Grok's training. Cursor, for its part, got access to SpaceX's Colossus supercomputer in Memphis -- roughly 200,000 Nvidia GPUs with plans to scale toward one million -- after publicly acknowledging it had been "bottlenecked by compute." "We've partnered with SpaceXAI to train Grok 4.5," Cursor's official account posted Wednesday. "It's our most powerful model yet and the first we've built for more than software engineering." SpaceX says the model reflects that pedigree: it "excels in large codebases and handles long-running tasks that span multiple repositories, hundreds of skills, and a variety of tools" -- precisely the messy, multi-file reality of professional software engineering that clean coding benchmarks often fail to capture. Early developer reactions suggest the training paid off. "Ok Grok 4.5 is wild," posted developer Evan Bacon. "It just built me this rocket tracking app with live data and a 3D globe. I might need a new benchmark after this." Inside xAI's turbulent year of scandals, departures, and rebuilding The polished launch belies how chaotic the road here has been. Grok has spent much of the past year in crisis. In mid-2025, the chatbot generated antisemitic content and at one point called itself "MechaHitler," episodes covered extensively by NPR and CNN. Earlier this year, its image-generation features allowed users to create sexualized deepfakes, including of children -- drawing investigations from the European Commission and Britain's Ofcom, as the BBC reported, and prompting SpaceX to list the behavior as a business risk in its own IPO filings. The organization behind the model was fracturing, too. All 11 of Musk's xAI co-founders had departed by the end of March, according to TechCrunch, and Musk publicly conceded that xAI "was not built right [the] first time around," saying he was rebuilding it "from the foundations up." Musk himself admitted at a conference this spring that Grok was "currently behind in coding" -- a rare public concession from an executive not known for them. Against that backdrop, Grok 4.5 reads as the first product of the rebuilt organization -- and the first proof point for the audacious story SpaceX told public market investors. During its IPO roadshow, the company pitched a total addressable market of roughly $28 trillion, with about $26 trillion tied to AI, including a $22.7 trillion "enterprise applications" opportunity. Those numbers strained credulity even by Silicon Valley standards. A competitive, cheap coding model is the most direct route from that narrative to actual revenue, which is why Wednesday's launch carries weight far beyond a routine model release. Grok 4.5 vs. Claude: the battle for the AI coding market The competitive stakes are hard to overstate, because the AI coding market has been consolidating around a single leader -- and it isn't Musk. Even as Cursor's revenue exploded, its market share was eroding. Spending data from Ramp cited by CNBC showed Cursor's share of the AI coding category falling from 41% in June 2025 to about 26% by May 2026, while Anthropic came to control roughly half the market. Anthropic also topped CNBC's Disruptor 50 list this year and, by Artificial Analysis's own measure, still holds the top spots on agentic performance rankings. That is the gap Grok 4.5 is engineered to close -- not by out-thinking Claude, but by underpricing it. The model's economics create a classic disruption dynamic: if it delivers most of the frontier's capability at a fraction of the cost per task, price-sensitive enterprise workloads will migrate, and incumbents will face pressure on their most profitable API traffic. The counterargument is that in coding, quality compounds. A model that resolves a complex bug correctly on the first attempt can be cheaper in practice than one that costs half as much per token but requires three tries. That is why Baker's caveat about "vibes" -- the developer community's shorthand for a model's felt reliability on real work -- will determine more than any launch-day benchmark. There is also a structural question buried in the deal. Cursor built its business on offering developers their choice of models, including Claude and GPT. If Grok becomes the favored child inside Cursor -- and Musk was already urging users to "Try out Grok 4.5 in Cursor!" within hours of launch -- the product risks alienating the very users whose data made Grok 4.5 possible. Regulators, already scrutinizing Grok on safety grounds in two jurisdictions, may take a keen interest in a company that controls the training data, the model, and a dominant distribution channel simultaneously. What Musk's trillion-dollar vertical integration bet means for AI's future Grok 4.5 also crystallizes what Musk's frenetic dealmaking was building toward. In February, SpaceX absorbed xAI in a share-exchange merger that CNBC confirmed valued the combined company at $1.25 trillion -- the largest merger of all time, valuing SpaceX at $1 trillion and xAI at $250 billion. The June IPO followed, the biggest in history, and the stock has since surged past $200 from its $135 offering price, vaulting SpaceX past Amazon and Microsoft to become the fourth most valuable company in the United States. The result is a single public company that owns nearly the entire stack: Colossus for training compute, ambitions for orbital data centers to power future scaling, a frontier model in Grok, a distribution channel in Cursor's developer base, and captive demand from Tesla and SpaceX's own engineering organizations. Neither OpenAI nor Anthropic can fully replicate that integration; both must reach developers through third-party tools, some of which Musk now owns. Whether that concentration proves to be an unassailable moat or a regulatory target -- or both -- is now one of the defining questions in enterprise AI. The next few weeks will start to answer it. Artificial Analysis says its full Intelligence Index results are forthcoming. Enterprise pilots will reveal whether the token-efficiency claims survive contact with real codebases. And Anthropic, which has answered every serious challenge this cycle with a rapid counter-release, is unlikely to cede the price-performance frontier quietly. But the deeper story of Grok 4.5 may be what it says about where the AI race has moved. For three years, the industry's scoreboard was intelligence: whose model was smartest. Musk, arriving late and battered, has chosen to compete on a different axis entirely -- whose model is cheapest to actually use. It is a telling choice from a man who built his fortune not by inventing the rocket or the electric car, but by relentlessly driving down the cost of making them. If the strategy works, Musk will have done to AI what he did to spaceflight. If it doesn't, he'll have spent $60 billion to learn that in software, unlike rockets, the cheapest ride isn't always the one engineers choose.

Elon Musk says 'Opus-class' Grok 4.5 is about to launch

Talk about jumping on the bandwagon at the very last minute. On Wednesday, SpaceXAI CEO Elon Musk announced that the new version of the company's AI chatbot, Grok 4.5, will be released to the public on Thursday, July 9. "Based on strong positive feedback from customers in our beta test program, @SpaceXAI will make Grok 4.5 available to the public tomorrow. It is an Opus-class model, but faster, more token-efficient and lower cost," wrote Musk in a post on X. The recently renamed AI/social networking/data-centers-in-space company is currently testing Grok 4.5 as private beta, while the stable version of the LLM is currently Grok 4.3. Musk calls the new Grok "Opus-class" in reference to the company's competitor Anthropic, whose general-purpose model, Claude Opus, is currently at version 4.6. Anthropic's most powerful model, Fable 5, has only recently been reinstated for users after being pulled following an intervention from the U.S. government. Meanwhile, OpenAI's powerful GPT-5.6 model is also launching on Thursday, following a short-lived launch ban from the White House.

SpaceXAI launches Grok 4.5 AI coding model for developers

SpaceXAI launched Grok 4.5, a new AI model built for coding and agentic tasks, priced at $2 per million input tokens and $6 per million output tokens. Access to the model is available across Grok Build, all Cursor plans, and the SpaceXAI developer console. Elon Musk, SpaceX's chief executive, described the model in a post on X $TWTR as "an Opus-class model, but faster, more token-efficient and lower cost," referencing Anthropic's top model family. By comparison, Anthropic charges $5 per million input tokens and $25 per million output tokens for Claude Opus 4.8, according to Reuters, while OpenAI's GPT-5.6 Luna comes in at $1 per million input tokens and $6 per million output tokens. SpaceXAI said the model runs at 80 tokens per second and delivers roughly twice the token efficiency of comparable leading models, meaning it completes tasks in fewer steps. The company said Grok 4.5 also supports finance and legal work, including building multi-sheet Excel models and creating PowerPoint presentations using native shapes. However, according to Axios, the model does not outperform the largest, most recent models from OpenAI and Anthropic. Grok 4.5 was trained alongside Cursor, the AI coding startup SpaceX agreed to acquire in a $60 billion all-stock deal expected to close in the third quarter of 2026. Cursor said in a statement that it partnered with SpaceXAI on the model's training. Development involved a collaborative exchange of data and compute between the two firms, with training conducted across tens of thousands of NVIDIA GB300 GPUs. European Union users will have to wait for access; SpaceXAI has indicated a mid-July target for that rollout. The launch comes during a period of heightened regulatory attention on powerful AI models. Following a government push to limit the rollout of GPT-5.6, OpenAI said the model would be released widely on Thursday, according to Bloomberg. The model also adds cybersecurity features; Cursor noted in its blog post that safeguards are in place to screen out malicious users without impeding legitimate vulnerability research. SpaceXAI was formed earlier this year after SpaceX folded xAI, Elon Musk's AI startup, into its operations.

The New Grok 4.5 Is Out. Elon Musk Says It Competes With Last Year's Claude Opus

Grok 4.5 is not available in the EU yet; SpaceXAI says European access is expected in mid-July. Elon Musk's SpaceXAI released Grok 4.5 on Wednesday, its first public model since the SpaceX-xAI merger closed in February and SpaceX's pending $60 billion deal to acquire Cursor. It targets coders, engineers, and what the company calls "knowledge workers" -- a category that apparently covers everyone from software developers to lawyers reviewing contracts to finance teams building Excel models. The company's pitch isn't that it's the best model. It's that it's cheap, for a western model at least. Grok 4.5 costs $2 per million input tokens and $6 per million output. Claude Opus 4.8, Anthropic's primary flagship, runs $5 input and $25 output. GPT 5.6 Sol, OpenAI's new top-tier model that also launched Wednesday, is priced at $5 input and $30 output. Musk posted on X and clarified where his new model actually sits. He called it "roughly comparable to Opus 4.7, but much faster." Opus 4.7 is Anthropic's previous flagship; Opus 4.8 has since succeeded it. Claude Fable 5 is now Anthropic's top of the line offering. He framed that as a deliberate tradeoff: speed and cost over raw capability, with engineers at Tesla and SpaceX as the proof of real-world utility. What the benchmarks actually show SpaceXAI published four benchmark results at launch, and the picture is mixed. DeepSWE 1.1 measures how reliably an AI can close real software bugs submitted by developers, using a standardized testing setup so models can be compared fairly, scored by percentage of issues fixed. Grok 4.5 scored 53%, behind Claude Opus 4.8 at 59% and GPT 5.5 at 67%. Claude Fable 5, Anthropic's frontier model, topped the chart at 70%. On SWE Bench Pro, another benchmark that measures a collection of software engineering problems scored by resolution rate, Grok 4.5 posted 64.7%, enough to beat GPT 5.5's 58.6% on that particular test. Opus 4.8 still leads at 69.2%, and Fable 5 sits at 80.4%. The company's benchmarks compare against GPT 5.5, not GPT 5.6, because the latter also launched Wednesday, hours after Grok 4.5's announcement. SpaceXAI trained Grok 4.5 in collaboration with the recently acquired Cursor AI on tens of thousands of Nvidia GB300 GPUs inside Colossus, the Memphis supercomputer with total capacity across more than 200,000 GPUs. The labs whose models sit above it on those same benchmarks don't own anything close to that hardware. The model that came out is competitive, just not first. That pattern has followed Grok across multiple releases. SpaceXAI has consistently turned up with enormous compute and third-place scores. What changed with Grok 4.5 is the pricing and the training signal. Where the case for it actually holds The better argument isn't raw performance; it's efficiency math. On SWE Bench Pro tasks, Grok 4.5 used an average of 15,954 output tokens to complete each job. Opus 4.8 burned through 67,020 tokens for the same work, a 4.2x gap. For teams running AI at volume, that difference compounds into real savings on top of the already-lower price per token. Even with Grok 4.5 scoring so low in the quality benchmarks, cheaper tokens and more efficiency in usage allow for more iterations without spending so much. The model also runs at 80 tokens per second, which is fast-model territory. Grok 4.5 was trained on developer session data from Cursor, including debugging traces and real code edits rather than static repositories. As Musk admitted in court, xAI's training practices have drawn scrutiny before; this time the pipeline runs through a platform SpaceX is in the process of buying outright. For developers running high-volume coding tasks, the math works: roughly Opus 4.7 capability at 60% less per input token. For anyone chasing the frontier, Claude Fable 5 leads every category SpaceXAI chose to publish. Our quick test using Grok build on Hermes was underwhelming for creative writing and acceptable on a simple coding task. The model is available via API, on Hermes and Grok build with half a million tokens of context (a little bit less than 400,000 words). European users will have to wait before using it; SpaceXAI says Grok 4.5 reaches the EU in mid-July.

Musk tells Tesla staff to switch to Grok -- a model he admits is worse

Elon Musk told Tesla staff to move to using Grok, the AI model from his own xAI (now folded into SpaceX), according to a memo sent to employees on Friday. The push comes days after Tesla capped employee spending on third-party AI tools -- and it lands even though Musk himself concedes Grok is not as good as its rivals. What the memo says Musk told staff they should switch to Grok "when possible" given Grok 4.5's lower token costs compared to competitors, according to the memo, first reported by The Information. He also asked engineers to email him directly with feedback on the model. The directive follows Tesla setting a $200 weekly limit on employees' AI spending earlier this week. That cap applies to models from Anthropic, OpenAI, and Google -- but pointedly exempts xAI's Grok, the tool Musk is now telling staff to adopt. Tesla has been testing beta versions of Grok internally for months, and xAI product lead Andrew Milich has been working with Tesla staff to troubleshoot issues. Despite that effort, four people familiar with internal usage previously said that Tesla engineers broadly prefer Anthropic's Claude for day-to-day development work. Grok 4.5 ranks below OpenAI, Anthropic, and Google xAI released Grok 4.5 on Wednesday, jointly with Cursor, the coding startup SpaceX is acquiring in a deal valuing it at $60 billion. Musk pitched the model as "Opus-class." The benchmark data tells a more modest story. On an aggregate multi-domain leaderboard, Grok 4.5 lands 9th overall at 76.3, behind multiple models from OpenAI (GPT-5.6 Sol, GPT-5.5, GPT-5.6 Terra, GPT-5.4), Anthropic (Claude Fable 5, Claude 4.8 Opus, Claude 4.7 Opus), and Google (Gemini 3.1 Pro). Its coding score of 68.6 is the lowest of any model on the board. The pattern repeats elsewhere. On LiveBench, the Grok family only just reached the bottom of the top tier -- the same spot occupied by open Chinese models that cost a fraction of the price. On the neutral DeepSWE 1.1 coding benchmark, which measures resolving real GitHub issues, Grok 4.5 scored 53% versus 70% for Claude Fable 5. There was also a benchmark problem at launch. Cursor disclosed, in a footnote, that an earlier snapshot of its own codebase was accidentally included in Grok 4.5's training data -- the very codebase its in-house benchmark tests against. That metric was excluded from the published comparison and the data removed for future models, but it inflated one of Grok's headline coding scores. Musk didn't dispute the gap. "In fairness, Fable is definitely better than Grok 4.5, but most tasks don't require Fable-level capability," he wrote on X. Grok 4.5's advantage is price: it runs at roughly $0.13 per task on the leaderboard above, versus $1.57 for Claude Fable 5. Electrek's Take Let's be clear about what's happening here. Tesla is not SpaceX, and it is not xAI, or SpaceXAI, or whatever Elon is calling his AI company this month. Tesla is a publicly traded company with its own shareholders and its own engineers. Those engineers shouldn't have their tools capped and then be steered onto a worse product simply because their CEO happens to own the company that makes it. That's the definition of self-dealing, and the $200 cap that conveniently exempts Grok makes the intent hard to miss. Now, I do get the underlying goal. AI coding tools are expensive, and reining in spending is reasonable. Grok 4.5 is genuinely cheaper. But it's cheaper because it's worse -- that's not a knock, it's the trade-off Musk himself just described. The problem is he's not letting engineers make that trade-off on the merits. He's mandating it, because when Tesla staff were left to choose, they kept picking Claude. And if cost is really the concern, there's a better answer than forcing everyone onto Grok: self-host. The open-weight field now delivers something like 90-95% of frontier capability at a fraction of the cost, and models like DeepSeek-V4 and GLM-5.2 are right there for the taking. Run them on hardware you already control and your marginal cost is power and amortized compute -- no per-token bill to anyone, including your CEO's other company. That would actually solve the spending problem. Telling engineers to use the boss's underperforming model doesn't -- it just moves the money to a different Musk entity. If you're managing rising energy costs at home the way Tesla is trying to manage its AI bill, home solar is one of the smartest ways to lock in low, predictable costs. With electricity rates climbing nearly 10% last year, home solar protects you against future rate increases. And with lease and PPA options, you can go solar with zero upfront cost and start saving immediately. If you want to find the best deal, check out EnergySage. It's a free service with hundreds of pre-vetted installers competing for your business, so you save 20 to 30% compared to going it alone. No sales calls until you pick an installer. Get your free quotes here.

SpaceXAI's newest AI model Grok 4.5 dramatically undercuts Anthropic and OpenAI on price

SpaceXAI's newest AI model Grok 4.5 dramatically undercuts Anthropic and OpenAI on price Elon Musk's SpaceXAI Corp. has released a new model called Grok 4.5, in what is its first major launch since it went public a few weeks earlier. In a blog post earlier today, the company said Grok 4.5 is designed to be a workhorse that's able to tackle all of the usual tasks that the artificial intelligence industry has been automating for some time already. That includes things like coding, writing emails and presentations, performing office and clerical work, doing research and other kinds of knowledge-based work. None of this sets Grok 4.5 apart, but SpaceXAI said that the difference is that it can do these tasks just as well as its peers at around half the cost, because it has "twice greater token efficiency" than its peers from other frontier model labs. If true, that could be a compelling advantage in a world where the cost of tokens has suddenly become a major issue for heavy AI users. SpaceXAI announced Grok 4.5's release alongside a host of benchmark results that highlight how competitive it is with the leading models of some of its major competitors, and it falls just short of their performance. In a post on the social media platform X, which is owned by SpaceXAI, founder Musk compared Grok 4.5 to Anthropic PBC's Opus, which is a large language model designed to handle intensive reasoning tasks. In a follow up, Musk added that the company's internal assessments show that Grok 4.5 is "roughly comparable" with Opus 4.7 in terms of its performance, but much faster at generating its results. "The combination of capability, faster speed and lower cost is what makes it competitive," he added. Grok 4.5's real calling card, however, appears to be its overall efficiency. The company said it costs around $2 per one million input tokens and $6 per one million output tokens, which makes it far cheaper than its rival's most capable models. In contrast, Opus 4.7 and 4.8 run at $5 per one million input tokens and $25 for one million outputs. Meanwhile, Fable 5, which is Anthropic's best model, costs $10 for inputs and $50 for outputs, based on one million tokens. OpenAI Group PBC, meanwhile, has a tiered pricing structure for different models. Its newest model, GPT-5.6 Sol, is priced at $5 for one million inputs and $30 for one million outputs, while Luna, its low cost version, costs $1 for one million inputs and $6 for one million outputs. This week is proving to be a big one in terms of new AI model launches. Earlier today, OpenAI announced the launch of GPT-5.6 Sol, its most powerful model so far, after being held up by the White House administration due to security concerns. According to OpenAI, GPT-5.6 Sol is its "strongest model yet," but those security concerns mean that it's currently only available to a limited number of customers. OpenAI also announced the launch of GPT-Live today, which is a family of AI models optimized to process spoken instructions.

Grok 4.5: SpaceXAI launches Grok 4.5 model for coding, agentic tasks

SpaceXAI on Wednesday launched the Grok 4.5 AI model, calling it the company's most intelligent offering to date designed for coding and agentic tasks. SpaceXAI on Wednesday launched the Grok 4.5 AI model, calling it the company's most intelligent offering to date designed for coding and agentic tasks. Here are some details: SpaceXAI said Grok 4.5 was trained across tens of thousands of Nvidia GB300 graphics processing units, with a focus on meticulous data filtering, deduplication and quality scoring. "We've ⁠partnered ⁠with SpaceXAI to train Grok 4.5," popular AI coding agent Cursor said. SpaceX said last month it would buy Anysphere, the startup behind Cursor, in an all-stock deal worth $60 billion to boost its presence in the lucrative enterprise AI tools market. Grok 4.5 is immediately available through SpaceXAI's AI coding agent, Grok Build, in Cursor and through the SpaceXAI ⁠console, the company's developer portal, using an API key. SpaceXAI said the EU availability is expected in mid-July. Grok 4.5 is ⁠priced at $2 per million input tokens and $6 per million output tokens, the company said. "It is an Opus-class model, but faster, more token-efficient and lower cost," SpaceX CEO Elon Musk said in a post on X. Musk's AI startup xAI was acquired by SpaceX in February. He said in May that xAI would cease to exist as a separate company and would instead become SpaceXAI. Rival Anthropic's Claude Opus 4.8 is priced at $5 per million input tokens and $25 per million output tokens. Comparatively, OpenAI's GPT-5.6 Luna is priced at $1 per ⁠million input tokens and $6 per million output tokens. Input tokens are the text, code or other data sent to an AI model, while output tokens are the text or code the model generates in response. OpenAI will publicly launch its most advanced AI model GPT-5.6 on Thursday, following a delay last month prompted by U.S. government requests over national security concerns about the potential misuse of powerful AI technologies.

Elon Musk's Grok 4.5 Could Rewrite Enterprise AI Economics - SpaceX (NASDAQ:SPCX)

On Wednesday, SpaceXAI launched Grok 4.5, its newest AI model designed to help users write code, complete complex work tasks, and handle research-heavy projects. SpaceXAI (formerly xAI) operates as a wholly owned artificial intelligence unit of Space Exploration Technologies Corp. (NASDAQ:SPCX). The company said Grok 4.5 is its strongest model so far and was trained alongside Cursor. SpaceXAI said the model can build apps from simple prompts, create Excel models, draft PowerPoint slides, and write clear documents in Word. The company priced Grok 4.5 at $2 per million input tokens and $6 per million output tokens, making it cheaper than some rival AI models. SpaceXAI's Grok 4.5 could pressure enterprise AI pricing by offering a lower-cost option for high-volume coding and agentic AI workloads, according to Counterpoint analyst Neil Shah. Grok Targets Enterprise AI Cost Pressure Shah said on Thursday that enterprises are facing "token bill shock" as autonomous agents and coding tools consume large volumes of tokens, making AI adoption increasingly expensive. He said Grok 4.5 enters the market as a fast, "good enough" and cheaper model priced at $2 per million input tokens and $6 per million output tokens, below Anthropic's Claude Opus 4.8 pricing of $5 for input and $25 for output. Analyst Sees Multi-Model AI Shift Shah said enterprises are moving toward diversified AI stacks, in which they route workloads based on cost, speed, and accuracy rather than relying on a single model provider. He said companies could use Claude for complex, high-stakes tasks while using Grok for high-volume developer workflows and repetitive agentic routing. Shah said Grok's access to Cursor telemetry data could help it improve through developer interaction feedback. He added that if Grok maintains its cost advantage while narrowing the accuracy gap, it could reshape enterprise AI economics and pose a new pricing threat to OpenAI and Anthropic. SPCX Price Action: SpaceX shares were up 0.88% at $149.60 during premarket trading on Thursday, according to Benzinga Pro data. Photo via Shutterstock Market News and Data brought to you by Benzinga APIs To add Benzinga News as your preferred source on Google, click here.

Elon Musk says Grok 4.5 to launch on July 9, calls it 'Opus-class'

xAI is speeding up development amid the race for the fastest and cheapest model in the AI space. Grok 4.5 first surfaced in private beta at Tesla and SpaceX in June, with Musk putting its early performance close to, and potentially above, Anthropic's Claude Opus. Elon Musk announced on X on Wednesday that xAI will release Grok 4.5, its latest large language model (LLM), to the public on July 9. "It is an Opus-class model, but faster, more token-efficient and lower cost," he wrote. xAI is speeding up development amid the race for the fastest and cheapest model in the AI space. Grok 4.5 first surfaced in private beta at Tesla and SpaceX in June, with Musk putting its early performance close to, and potentially above, Anthropic's Claude Opus. None of Musk's performance claims have been independently benchmarked yet; they are based on internal testing at Tesla and SpaceX rather than public evaluations. Grok 4.5 runs on a new 1.5-trillion-parameter foundation model known as V9, with supplemental training on data from the AI coding platform Cursor to sharpen its technical and coding chops. SpaceX is in the process of acquiring Cursor in a $60 billion deal. The release comes under the SpaceXAI banner, the name xAI being adopted after the organisation was folded into SpaceX earlier this year. The latest model would follow Grok 4, launched in July 2025. The company has mapped out a roadmap of monthly model releases through the rest of 2026, Musk had said in June. The release puts Grok 4.5 on a collision course with OpenAI, which is moving its GPT-5.6 models toward broader availability around the same period. This sets up another head-to-head moment in a race already crowded with players such as Claude, Gemini, and ChatGPT.

SpaceX Readies Expansion of Grok AI Model | PYMNTS.com

Grok 4.5 will become available to the public Thursday (July 9), SpaceX CEO Elon Musk wrote in a post on the company-owned social media platform X early Wednesday (July 8). "It is an Opus-class model, but faster, more token-efficient and lower cost," Musk wrote. The trillionaire had announced late last month that the model had entered private beta testing at SpaceX and Tesla, saying it could rival Anthropic's Claude Opus. A report Wednesday on the rollout from Seeking Alpha, citing Apptopia data, notes that the number of average daily users for the Grok app has fallen 28% since April. The app's market share was at 8.7% last month, down from 10.6% in May, the report added, while OpenAI's ChatGPT held the top ranking among American market share in June. Meanwhile, the Wall Street Journal reported last week that SpaceX had developed a handset-style AI device and shown the prototype to investors before going public. The device, said to be slimmer than an iPhone, was built to run on a proprietary operating system and integrates AI technology from SpaceX's xAI, sources told the news outlet. Musk later denied the report, calling it "utterly false" in a post on X, leading to a drop in SpaceX's share price. PYMNTS wrote last week about the potential behind such a device, pointing out that SpaceX controls Starlink, a satellite network with global coverage, and is building direct-to-cell service that allows phones to connect to satellites without a terrestrial carrier. And SpaceX's merger with xAI in February, a deal said to value the combined company at around $1.25 trillion, gives it direct access to Grok. "A consumer device connecting to all three would give SpaceX a hardware endpoint that bypasses both the app stores and cellular carriers," the report said, adding that no existing AI advice offers that combination. "SpaceX went public in June in a record-setting initial public offering (IPO)," the report added. "A public company's investor narrative carries different weight than a private one. A phone-shaped slot in the IPO pitch deck, with xAI models inside and Qualcomm silicon underneath, frames SpaceX as a vertical platform company rather than a launch and connectivity provider." The report also noted the uneven track record of AI devices. "From Google Glass to the Humane AI Pin, the pattern is consistent. Consumers don't adopt hardware because it's futuristic," PYMNTS wrote. "They adopt it because it solves a problem better than what they already carry.

Elon Musk pulls no punches with AI rivals as Grok 4.5 debuts

The best product doesn't always win. In most markets, the winner is the product that's good enough at a price nobody can ignore. Toyota understood that. Southwest Airlines built an empire on it. Now the same playbook is being tested in the most expensive technology race in history. Artificial intelligence labs have spent the past year one-upping each other on capability, and businesses have paid for it. Every automated coding agent and research assistant runs on tokens, the units of text AI models read and write, and monthly bills have grown so fast that some companies now treat AI spending like a second cloud budget. That tension is the backdrop for the newest move from Elon Musk, whose rocket company became a publicly traded AI bet with its June 12 initial public offering (IPO) and has spent the weeks since trying to convince Wall Street that the second half of that description is real. On Wednesday, July 8, SpaceXAI, the artificial intelligence unit of SpaceX (SPCX), launched Grok 4.5, and the sales pitch is unlike anything Musk has tried before. He isn't claiming he built the best model in the world. He's claiming he built the one you can actually afford to run all day. What Grok 4.5 brings to the AI fight The model was "trained alongside Cursor" and built for coding, agentic tasks, and everyday knowledge work, according to SpaceXAI. Agentic tasks are jobs an AI finishes on its own across multiple steps, such as finding a bug, fixing it, and testing the result. That Cursor reference matters. SpaceX agreed in June to buy Anysphere, the startup behind the Cursor coding tool, in a $60 billion all-stock deal, as TheStreet covered, and Grok 4.5 is the first model to emerge from that pairing. The training run used tens of thousands of Nvidia (NVDA) GB300 graphics processing units, according to SpaceXAI. Grok 4.5 runs on the company's new V9 foundation model with roughly 1.5 trillion parameters, about three times the size of its predecessor, Musk said on X. Grok 4.5 arrived roughly three months after Grok 4.3 shipped in April, following weeks of private beta testing inside SpaceX and Tesla, Musk wrote earlier this month. Developers got access on July 8 through Cursor, the Grok Build coding agent and the SpaceXAI developer console, with the public rollout following on Thursday, July 9. European availability is expected in mid-July, and usage is free for a limited time in Grok Build and Cursor, the company confirmed. The company's own benchmark charts tell an honest story. Grok 4.5 topped rivals on the SWE Marathon software engineering test at 29%, but trailed Anthropic's Claude Opus 4.8 and Claude Fable 5 on SWE Bench Pro, scoring 64.7% against their 69.2% and 80.4%, according to SpaceXAI's published figures. SOPA Images / Getty Images Musk's pricing play targets OpenAI and Anthropic The launch was never really about benchmarks. It was about the invoice, and Musk made sure everyone knew it. Here is how the launch pricing compares per million tokens: * Grok 4.5 costs $2 for input and $6 for output, SpaceXAI noted. * Anthropic's Claude Opus 4.8 costs $5 for input and $25 for output, according to Reuters. * OpenAI's GPT-5.6 Luna costs $1 for input and $6 for output, Reuters added. Musk framed the tradeoff himself. "It is an Opus-class model, but faster, more token-efficient and lower cost," the SpaceX CEO said in a post on X, according to Reuters. Then he went further than most executives would. "In fairness, Fable is definitely better than Grok 4.5," Musk wrote of Anthropic's flagship model, adding that most tasks don't require that level of capability, as reported by Stocktwits. I ran the numbers on what that gap means in practice. A company generating one billion output tokens a month, a realistic volume for a mid-sized engineering team running coding agents, would pay about $6,000 on Grok 4.5 versus roughly $25,000 on Opus 4.8. Across a year, the difference approaches a junior developer's salary. That is the emotional core of this launch for anyone who signs an AI invoice. Musk isn't selling brilliance. He's selling relief. What the Grok 4.5 gamble means for SPCX investors The market's first reaction was a shrug. SpaceX shares fell nearly 1% on Wednesday, July 8, to $148.30, a third straight decline that left the stock down about 8% for the week, Stocktwits reported. Shares edged up 0.88% to $149.60 in the premarket of July 9, according to Benzinga. More Artificial Intelligence: Some context helps here. SpaceX priced its IPO at $135 a share, raised a record $75 billion in the June 12 debut, and briefly traded above $176 before giving nearly all of those gains back. The muted response fits a stock that has spent its first month erasing its post-IPO pop. Jim Cramer has already warned buyers about the one-way momentum, as seen in my TheStreet coverage. Morningstar said before the debut that it doesn't count Grok among the leading AI labs, TheStreet highlighted. Pricing is the variable that could change that conversation. Enterprises running autonomous agents are facing "token bill shock," Counterpoint Research analyst Neil Shah said July 9, Benzinga confirmed. If Grok holds its cost advantage while narrowing the accuracy gap, it could squeeze pricing at OpenAI and Anthropic, Shah added. My analysis is that Musk picked the one fight he can win right now. He can't out-benchmark Anthropic this quarter, and he admitted as much in public. He can out-price it, however. Musk also told users to expect noticeable gains in the Grok Build coding harness every week, a promise that shifts the story to shipping cadence rather than one launch-day scoreboard. The timing adds pressure, too. OpenAI's GPT-5.6 arrives Thursday, July 9, which means the cost-versus-capability debate will sit at the center of the AI trade for the rest of the summer. Wall Street remains constructive, despite the wobbly chart. Analysts hold a strong buy consensus on SpaceX with an average price target of $212.08, implying roughly 40% upside, according to TipRanks. For investors, the test is no longer whether Musk can build a frontier model. It's whether "good enough" at a quarter of the output price shows up in SpaceXAI revenue before the lockups expire and Wall Street's patience runs out. The Arena Media Brands, LLC THESTREET is a registered trademark of TheStreet, Inc. This story was originally published July 10, 2026 at 10:17 AM.

SpaceXAI launches Grok 4.5 AI model for coding, agentic tasks and knowledge work

SpaceXAI has announced Grok 4.5, its latest artificial intelligence model designed for coding, agentic tasks, and knowledge work. The company describes it as its most advanced model to date and said it was developed alongside Cursor. Grok 4.5 According to SpaceXAI, Grok 4.5 was trained on datasets covering coding, science, engineering, and mathematics using tens of thousands of NVIDIA GB300 GPUs. The training process included data filtering, deduplication, quality scoring, domain-specific data selection, and stability techniques designed for large-scale runs. The company also expanded reinforcement learning with hundreds of thousands of multi-step software engineering and technical tasks using automated and model-based grading. SpaceXAI said its asynchronous training infrastructure enables long-running agentic rollouts while training continues across tens of thousands of GPUs, improving reasoning on engineering and agentic tasks. The company also claims Grok 4.5 outperforms comparable leading models on real-world engineering tasks. Coding and productivity capabilities Grok 4.5 supports software development across languages including Rust and C/C++, and can generate complete applications from a single prompt. SpaceXAI said the model delivers inference speeds of up to 80 tokens per second (TPS) and achieves roughly 2x better token efficiency than comparable leading models, allowing it to complete tasks using fewer generated tokens. The company added that this enables Grok 4.5 to deliver higher intelligence per unit of time and cost. Grok 4.5 is also the default model in Grok Build, where it can: * Build complex Excel workbooks involving web research and multi-sheet formulas * Leave reference notes within Excel files * Generate Microsoft Word documents * Create PowerPoint presentations and diagrams using native PowerPoint shapes Pricing and availability Grok 4.5 is priced at: * $2 per million input tokens * $6 per million output tokens The model is available through Grok Build, Cursor on all plans, and the SpaceXAI console. SpaceXAI is also offering free Grok 4.5 usage for a limited time in Grok Build and Cursor. The company noted that Grok 4.5 is not yet available in the European Union across its products or API console, with availability expected in mid-July.

SpaceXAI Launches a Frontier Model Just as OpenAI Gets Its Premium GPT-5.6 Ready to Roll

Did it ever occur to you dear reader that the AI battle has become something of an ego tussle where customers are too hardly worth caring about? How else does one explain the fact that SpaceX's AI unit is rolling out its Grok 4.5 model for coding and agentic tasks hours after OpenAI gets the Trump administration sign-off to roll out its most advanced GPT-5.6 version. Just so readers are aware, Sam Altman isn't one to take this lying down. His OpenAI launched the GPT-Live, their new generation of voice models for natural human-AI interaction as a direct counter to SpaceX's voice agent builder launched a week ago. Maybe it is nothing more than a coincidence, but given the palpable friction between Elon Musk and Sam Altman, we cannot be blamed if conspiracy theories flash within our heads. In recent times the SpaceX trillionaire has showed a marked soft-corner for Anthropic, even giving away excess AI infrastructure capacity to the Claude-maker - albeit for a fee. In fact, while introducing Grok 4.5 jointly with soon-to-be-subsidiary Cursor, Musk promoted the release via his X handle as "roughly comparable to Opus 4.7, but much faster." And here we thought that GPT-5.5 was doing a better job with coding, agentic tasks and knowledge work - the three things that SpaceXAI claimed their frontier model excelled at. But that wasn't all. He took potshots at OpenAI's penchant for boasting about beating benchmarks in the post. "The combination of capability, faster speed and lower cost is what makes it competitive. We are closing the loop on real-world usefulness, not benchmarks. Hardcore engineers at Tesla & SpaceX find Grok 4.5 genuinely useful, which is what actually matters." Touché? The other reason we are confused about Musk's real motives is that just last month he was busy leasing out excess AI infrastructure capacity, something that detractors of SpaceX had called its biggest financial drags. Out of nowhere, Musk signed deals with Anthropic and Google to share capacity on the Colossus 1 datacentre, netting $26 billion in annualised compute revenue. The two deals happened between the first week of May and the first week of June. In July, we have the company announcing AI products of value to businesses. A week ago, SpaceX had introduced a Voice Agent Builder that allows small business to set up an AI that can talk on the phone and handle customer services calls. So, what exactly happened between these three distinct periods of time? We scoured the internet and found nothing exciting other than President Trump's crackdown on AI models that could cause national security risks. First, he ordered Anthropic to pull out Claude Mythos 5 from global markets and then asked OpenAI to hold off its GPT-5.6 from a global release. There was consternation around Trump's moves both domestically and globally with several countries including India quietly seeking options beyond the US AI industry. However, things normalised as Mythos 5 and Fable 5 (a guard railed version) were allowed to export and on Tuesday OpenAI was also given the go-ahead to rollout GPT-5.6. While this theatre of the absurd was playing out elsewhere, what does Elon Musk do? He ratchets up the absurdity levels by launching two solutions that is quite obviously late to the party and could only convince his fanboys SpaceX being serious about AI. The voice agent comes into a market already inhabited by the likes of Sierra while a coding model is hardly new. Just so readers can draw their own inferences, OpenAI will launch its most capable GPT-5.6 Sol along with the lower-cost Terra and Luna models later today. Altman's team touted the improved agentic capabilities in coding, biology and cybersecurity and placed itself as a competitor to Anthropic's Mythos Preview. And then what happens? OpenAI last night launched GPT-Live, its "new generation of voice models for natural human AI interaction. It was rolled out to all ChatGPT users on Go, Plus, and Pro plans. Free user rollout is in progress, the company said on its X handle - right under the very nose of Elon Musk with whom OpenAI is also having a prolonged legal battle. Who in their right minds wouldn't think that all of the above were just coincidences? In fact, if anything the manner in which Musk has gone about his latest riposte to OpenAI makes us believe that we are right with our theories. For starters, SpaceX made the latest announcements without much fanfare - something that is anathema to Musk's need for showmanship. Then there is the small issue of Musk using only his X account to promote the events. Does he assume that small businesses, their target audience, actually spend time online waiting for Musk's pearls of wisdom?

SpaceXAI Launches Grok 4.5 Ahead of GPT-5.6 Race: What We Know So Far

SpaceXAI launches Grok 4.5 as the AI race heats up, with Elon Musk aiming to challenge top rivals while bringing smarter AI features to millions of X users. The AI race has become even more intense. SpaceXAI will launch Grok 4.5, its latest language model, on July 9, 2026. The new model arrives before the expected release of next-generation AI systems from several rivals, making the timing hard to ignore. Elon Musk has been clear about his goal. He wants Grok to stand alongside the best AI models available today. With every new release, the company is trying to improve reasoning, speed, and the overall user experience. Grok 4.5 is another step in that plan.

SpaceX to make Grok 4.5 available to public tomorrow, Musk says By Investing.com

Investing.com-- SpaceX will make its Grok 4.5 artificial intelligence model available to the public tomorrow, CEO Elon Musk said in a social media post on Tuesday evening. Musk described the model as "Opus-class", comparing it to Anthropic's Claude, but said it was "faster, more token-efficient and lower cost." Get more breaking news on the biggest AI stocks by subscribing to InvestingPro Grok is the flagship AI model of Musk's xAI, which was folded into SpaceX this year and recently renamed to SpaceXAI. XAI had last released Grok 4.3 in April, with Musk having long teased the next version of his flagship AI. Musk had earlier in July said Grok 4.5 had entered private beta testing at SpaceX and Tesla, and that the model was built on xAI's new 1.5 trillion-parameter V9 foundation model.

Grok 4.5 and Cursor were trained together: It could change who codes with what

This week, SpaceXAI, created by Elon Musk, has released its most important update ever since the company has agreed to purchase Cursor for 60 billion dollars back in June. The most impressive thing about Grok 4.5 is that it has been trained using Cursor's dataset which consists of trillions of tokens collected from the actual users. These include code bases as well as interactions between developers and agents. It is quite a different release from what we have seen before. Also read: Full Duplex Voice explained: What it is and how ChatGPT's voice has evolved There are some aspects about this release that you need to know to understand the significance of it. In June, SpaceXAI has announced that it will purchase an AI coding startup called Cursor for 60 billion dollars. The startup itself has been developing an in-house model which is specifically designed for coding - Composer 2.5. But Grok 4.5 goes way further than that. It includes STEM and research paper tasks together with all the Cursor data which allows the model to cover such areas as finance, law and general tasks. According to Elon Musk, "it's an Opus-class model, but faster, more token-efficient and lower cost". Does this signify Grok making its way into the mainstream of vibe coding? It doesn't, at least not yet, and the narrative can use some challenging here. Grok 4.5 is not taking any place from GPT-5.6, Opus 4.8, and Cursor's own Composer 2.5 in the pool; rather, it receives default visibility: double usage limits in the first week, availability on desktop, web, iOS, CLI, and the SDK, so millions of real-life developers who don't necessarily seek Grok 4.5 out will encounter it. This might be distribution, but it's often the first step toward mainstream. Also read: We tested every claim Meta made about Muse Image: Here's what held up It should be noted right away that there is a catch regarding the benchmark charts. An accidental inclusion of the previous version of the codebase of the same company in the training dataset of Grok 4.5 made the score on the internal benchmark of Cursor significantly higher, and the company decided to exclude the metric altogether from public comparison. Moreover, some of the metrics used to demonstrate superiority to the competition are reported by the company itself and/or obtained in the internal testing of Cursor. None of that erases what's genuinely interesting here, but it does mean the "beats Opus 4.8" framing floating around deserves a raised eyebrow rather than a retweet. The more durable story is structural. SpaceXAI trained this model partly on the same compute it rents out to Anthropic and Google for close to a billion dollars a month, and now owns the IDE where a meaningful chunk of professional coding actually happens. Musk has built himself a data flywheel: buy the tool, train on the usage, ship the model back into the tool. Whether Grok 4.5 wins developers on merit is still an open question. Whether Musk has built the pipes to keep trying is not.

SpaceXAI launches Grok 4.5, claims major gains in coding, engineering and enterprise tasks

Grok 4.5 is available now on Grok Build, Cursor, and the SpaceXAI API, with EU rollout expected later this month. Elon Musk's SpaceXAI has introduced Grok 4.5, the latest version of its AI model. It comes with improvements in coding, software engineering and enterprise productivity. The company says the new model is its most capable AI system yet and has been designed to tackle complex, multi-step tasks with faster reasoning while using fewer computing resources. As per SpaceXAI, Grok 4.5 has been trained on large scale datasets covering programming, mathematics, science and engineering. The company says it has improved the training pipeline by improving data quality, filtering duplicate information and using reinforcement learning techniques focused on real-world software development. The company claims Grok 4.5 performs strongly across several industry benchmarks used to evaluate coding and engineering capabilities. On the SWE Marathon benchmark, which measures performance on long-running software engineering tasks, Grok 4.5 reportedly outperformed competing AI models. It also posted competitive results on benchmarks including DeepSWE, Terminal Bench, and SWE Bench Pro, though some rival models continued to lead in certain categories. In a blog post, SpaceXAI also stated that the model requires considerably fewer output tokens to complete programming tasks than some competing AI models, allowing it to generate responses faster while reducing inference costs. The company also claims the model delivers responses at speeds of up to 80 tokens per second. Along with the software development, Grok 4.5 is also positioned as a productivity tool for enterprise users. It can generate Excel spreadsheets with formulas, create PowerPoint presentations using native design elements, draft Word documents and assist with research-intensive office work. The model is now the default AI engine inside Grok Build, where it can build applications from a single prompt and assist with complex document creation. SpaceXAI has priced Grok 4.5 at $2 per million input tokens and $6 per million output tokens. It is available starting today through Grok Build, Cursor, and the SpaceXAI API. However, the rollout excludes the European Union for now, with availability there expected later this month.