SpaceXAI Launches Grok 4.6 AI Model, Matching OpenAI's GPT-5.6 Sol on Key Benchmarks

Grok Gets Cursor-Driven Upgrade, Claims to Be Competitive With Top Models

Thus far in its short lifespan, the AI model Grok is probably best known for being used to mass-produce non-consensual nude images and for declaring itself MechaHitler. So it's an uphill battle to convince people to use it for coding and knowledge work. But for those heavily invested in keeping tabs on the benchmarks of AI models, the release of Grok 4.6, which was made publicly available on Wednesday, might be compelling. According to SpaceX's xAI, the new model "achieves frontier intelligence across several agentic coding and knowledge work benchmarks." The company also claims that Grok 4.6 matches OpenAI's GPT-5.6 Sol on a number of benchmarks, including the Artificial Analysis Intelligence Index, which is a composite score of nine benchmarks. The marks are high enough for CEO Elon Musk to dig into the hyperbole playbook. On X (also owned by SpaceX), Musk claimed that Grok 4.6 is "objectively #1 when considering intelligence, speed & cost" -- which doesn't seem like the kind of thing that you'd say when the biggest brag in your announcement is that your model matched (as opposed to beat) a competitor, but it's about in line with Musk's typical brand of braggadocio. Grok's return to the frontier race after lagging behind competitors like OpenAI and Anthropic for quite some time comes after the company partnered with and maybe acquired agentic coding company Cursor. Grok 4.5 was the first version of xAI's model to be trained on, in part, the trove of real-world usage data that Cursor has accumulated, and Grok 4.6 reportedly underwent an even longer supplemental training run than its predecessor. Given that, it seems possible that Musk's seemingly panicked acquisition of Cursor right before SpaceX went public may have, at least temporarily, lifted Grok from potentially falling into obscurity in the AI race to being a player. The company is certainly leaning into the Cursor of it all. Grok 4.6 will first be available in Cursor, along with Grok Build, the xAI coding agent, starting immediately. Cursor and xAI also teamed up on a persistent, always-on AI agent called Grok Bot, which it released in beta earlier this week. That doesn't mean Grok won't have some work to do to flip its reputation. On top of the whole controversy in which the model was used to generate massive amounts of non-consensual nude images (including of children), the company hasn't even really been able to win major allies. The second Trump Administration, which Musk spent $400 million to make a reality, hasn't shown much interest in using Grok -- and multiple agencies have even raised concerns about its safety. And Grok is basically a non-entity in enterprise at this point, with just 4% of companies that have adopted AI tools choosing to pay for xAI's option, according to Ramp's AI Index. It's theoretically good for an AI model to have a niche, but Grok's current one is probably not going to be great for business. Cursor's team seems to have been able to help turn around Grok's performance, but the reputation part will likely require some work.

SpaceXAI debuts Grok 4.6, overtaking Kimi K3's performance and vaulting to the world's fourth best on Artificial Analysis

Elon Musk's company SpaceXAI, formerly known as xAI, has released Grok 4.6, its latest frontier AI model, with a focus on long-running agents, coding and knowledge work -- and a pricing strategy designed to make those workloads cheaper to run. The model scores 61 on the third-party Artificial Analysis Intelligence Index, surpassing the popular open weights Chinese model from Moonshot, Kimi K3, and tying rival OpenAI's GPT-5.6 Sol Max and improving five points over Grok 4.5 High. Anthropic's Claude Opus 5 and Fable 5 occupy the number one and two spots, respectively. More consequential for enterprises evaluating AI agents, Grok 4.6 posts sizable gains over its predecessor across coding, terminal, knowledge-work and agent benchmarks while retaining an application programming interface (API) price starting at $2 per million input tokens and $6 per million output tokens, making it a mid-priced frontier model comparing leading options that are both proprietary and open source, globally, according to VentureBeat's analysis. Still, that's less than half of what GPT-5.6 Sol costs over OpenAI's API in standard mode. SpaceXAI says Grok 4.6 is available today in Grok Build, SpaceXAI's answer to Anthropic's Claude Code and OpenAI's Codex, which is available starting in the $30 per month SuperGrok plan. It's also available in SpaceX's recent acquisition of the AI coding startup Cursor, and from partners including OpenRouter, Vercel and Cloudflare. SpaceXAI is providing twice the included usage for Grok 4.6 in Cursor and Grok Build during the first week. The release arrives only weeks after Grok 4.5, which SpaceXAI launched in July as a model targeting coding, agentic tasks and knowledge work, and one day after the launch of Grok Bot, a new system for assigning AI agents to complete designated tasks as virtual employees. The bigger change is agent behavior, not just another benchmark point SpaceXAI describes Grok 4.6 as being built specifically to stay on task across longer sequences of work, including researching unfamiliar topics, analyzing information, navigating codebases and converting product ideas into working applications. The company says it subjected the model to a longer supplemental training run than Grok 4.5, using curated model-generated reasoning and technical data alongside engineering data and changes to its optimizer and training recipe. It then used Grok 4.5 to regenerate supervised fine-tuning trajectories across reasoning levels, agent harnesses, STEM, software engineering and knowledge work, filtering problematic trajectories with model-based checks. Reinforcement learning also targeted agentic environments spanning general coding, knowledge work, kernel optimization, web development and computer-aided design. That matters because enterprise AI deployments are increasingly moving beyond isolated prompt-and-response interactions toward agents expected to maintain state, operate tools, modify code and recover from problems across longer execution paths. SpaceXAI says that during its testing, Grok 4.6 showed more self-testing and verification on longer trajectories, checking its own work before proceeding. It also reports stronger first attempts on interactive and visual projects than Grok 4.5. Those are company observations rather than independent guarantees of production behavior, but they indicate where SpaceXAI concentrated the model's post-training work. Grok 4.6 reaches the frontier, but does not sweep it Grok 4.6's improvement over Grok 4.5 at this juncture of the AI model competition cannot be overstated. According to Artificial Analysis, Grok 4.6 reaches an Elo score (human preference of head-to-head model outputs, adapted from chess) of 1,753 on GDPVal-AA v2, the benchmark measuring performance on real-world tasks like scheduling and diagramming, versus 1,526 for Grok 4.5, 1,728 for GPT-5.6 Sol Max and 1,741 for Fable 5 Max. The coding results from SpaceXAI show a similar generational improvement but more competition at the frontier. Grok 4.6 scores 69.9% on CursorBench v3.2, up from 66.7%, while Fable 5 Max reaches 70.5%. On DeepSWE v1.1, Grok rises sharply from 54% to 65.9%, but GPT-5.6 Sol Max leads at 73%. FrontierCode v1.1 Extended moves from 56.6% to 61.3%, compared with 60.6% for GPT-5.6 Sol Max and a leading 63.6% for Fable 5 Max. Agent benchmarks tell much the same story. Grok 4.6 reaches 57.5% on APEX-Agents, a 10.4-point increase over Grok 4.5's 47.1%, narrowly exceeding GPT-5.6 Sol Max's 56.7% but trailing Fable 5 Max at 59.2%. On APEX-SWE, Grok 4.6 rises to 56.4% from 53.6%, while Fable 5 Max scores 58.8%. Terminal-Bench v3.0 exposes a larger remaining gap. Grok 4.6 improves from 15.7% to 26%, but GPT-5.6 Sol Max and Fable 5 Max score 34.6% and 34.1%, respectively. Two of Grok 4.6's strongest results come from longer-horizon professional work. On AA-Briefcase it scores an Elo of 1,577, narrowly exceeding Fable 5 Max's 1,574 and topping GPT-5.6 Sol Max's 1,502. On Harvey LAB, Grok 4.6 reaches 15.8%, versus 12.9% for Grok 4.5, 11.3% for Fable 5 Max and 2.5% for GPT-5.6 Sol Max. SpaceXAI notes an important methodological caveat: third-party scores in its table use the best self-reported or publicly available results. The comparison therefore should not be interpreted as a perfectly controlled four-model evaluation. In other words, the evidence supports a substantial upgrade over Grok 4.5 more clearly than it supports across-the-board superiority over rival frontier models. Grok 4.6 wins several of the displayed evaluations while GPT-5.6 Sol Max and Fable 5 Max retain meaningful leads elsewhere. Cost could be the more important enterprise benchmark Artificial Analysis' supplied evaluation adds another dimension: how much work the model performs for the money spent. The testing places Grok 4.6 on its Intelligence-versus-Cost-per-Task Pareto frontier at a reported $0.84 per task -- which actually makes it less of a bargain than its predecessor, Grok 4.5, and less economical than OpenAI's GPT-5.6 Luna, z.ai's GLM-5.2, and Meta's new Muse Spark 1.2, among other models. Artificial Analysis also reports that Grok 4.6 completed its AA-Briefcase workloads in roughly 53 turns and about 0.5 billion input tokens on average, versus approximately 103 turns and 2 billion input tokens for Claude Opus 5 Max. Those measurements do not prove that every production agent will use fewer tokens or finish twice as quickly. Agent costs depend heavily on harness design, prompts, tool calls, caching, retry behavior and the task itself. But they point toward an increasingly important enterprise metric: the cost of completing a workflow, rather than simply the cost of generating one million tokens. That distinction is central to SpaceXAI's positioning. The standard Grok 4.6 API starts at $2 per million input tokens and $6 per million output tokens, and SpaceXAI also offers a faster variant at twice the price. The supplied API documentation adds an important caveat for long-context deployments. Grok 4.6 supports a 500,000-token context window, but prompts below 200,000 tokens are billed at $2 per million input tokens, $0.50 per million cached-input tokens and $6 per million output tokens. Once a prompt reaches 200,000 tokens, those rates rise to $4, $1 and $12 respectively, with the higher pricing applying to all tokens in that request. That means enterprises should not extrapolate the $2/$6 headline pricing across the model's entire context window when estimating total cost of ownership. Artificial Analysis says the standard headline rates remain more than 60% below the competing frontier-model prices it cites for Claude Opus 5 and GPT-5.6 Sol. The practical savings will depend on how many tokens each model consumes to complete the same workload. The Grok name carries considerable baggage and controversy, separate from the general AI skepticism Performance and price may not be the only hurdles SpaceXAI faces in converting Grok 4.6's benchmark gains into enterprise adoption. The Grok brand arrives with an unusually visible history of safety and governance controversies -- including extremist and antisemitic outputs, politically skewed responses, exaggerated praise of Elon Musk and, more recently, the use of Grok's image-generation capabilities to produce non-consensual sexualized imagery. For companies with strict compliance, brand-safety or responsible-AI requirements, that history could become a procurement consideration separate from the technical capabilities of Grok 4.6 itself. The most notorious text-generation episode came in July 2025, when Grok produced antisemitic posts, praised Adolf Hitler and in some responses referred to itself as "MechaHitler." SpaceXAI's predecessor xAI subsequently said it was removing inappropriate posts and taking steps to prevent hate speech from being published by Grok. Also in summer 2025, Grok began inserting references to an alleged "white genocide" in South Africa into answers to unrelated questions. xAI said an unauthorized modification to Grok's response software had directed the system to produce a particular response on a political topic while bypassing its normal review process. The company said the change violated its policies and subsequently pledged to publish Grok's system prompts and establish round-the-clock monitoring for problematic responses. The South African government has rejected claims that a genocide against white South Africans is taking place. Grok's objectivity came under scrutiny again in November of the same year after the chatbot repeatedly produced implausibly flattering assessments of Musk. Among the examples reported at the time were claims placing Musk above elite athletes and historic intellectual figures. Musk said Grok had been manipulated through adversarial prompting into making "absurdly positive" statements about him. Whatever the underlying cause, the incident illustrated the reputational problem for an enterprise model whose outputs can become entangled with the public persona of the executive most closely associated with its developer. The most serious controversy has involved image generation. In January 2026, U.K. regulator Ofcom opened a formal investigation into X after reports that the Grok account was being used to create and distribute undressed images of people and sexualized images of children. Ofcom said the material under examination could amount to non-consensual intimate-image abuse, pornography and child sexual abuse material. X subsequently said it had implemented measures intended to stop the Grok account from being used to create intimate images of people, but Ofcom said its investigation remained open. The scrutiny extends beyond Ofcom. Britain's Information Commissioner's Office is investigating X and xAI over both the development and deployment of Grok, including whether personal data was handled lawfully and whether adequate safeguards existed to prevent harmful manipulated imagery. The European Commission, meanwhile, opened a separate formal investigation under the Digital Services Act examining X's management of systemic risks connected to Grok, including the dissemination of manipulated sexually explicit material. Those investigations concern X and the earlier xAI organization rather than establishing a finding that the newly released Grok 4.6 API violates those laws. Nevertheless, they are unlikely to help SpaceXAI sell Grok to businesses. SpaceX acquired xAI in February 2026, and the AI operation now markets itself as SpaceXAI, meaning Grok's newest models sit under a different corporate structure but retain the same consumer-facing brand. There is no evidence in the material examined here that Grok 4.6 itself repeats the specific "MechaHitler," "white genocide," sexual-image or Musk-flattery incidents associated with earlier Grok deployments. But enterprise procurement teams rarely evaluate a model in isolation from its vendor and product history. For SpaceXAI, that means Grok 4.6 may have to demonstrate not only that it is cheaper or more capable than competing frontier models, but that the controls around it are sufficiently predictable for organizations that cannot afford their AI supplier to become a brand-safety event. That continuity creates a potential adoption problem that benchmark tables cannot measure. Developers choosing a model for an internal coding agent may care primarily about price, latency and task completion. A bank, government agency, healthcare provider or consumer brand deploying the same model into customer-facing or regulated workflows may also have to consider vendor governance, content-safety controls, auditability and reputational exposure. A model designed to be deployed, not just chatted with Grok 4.6 supports text and image inputs with text output, function calling, structured outputs and reasoning, according to the supplied API specifications. Those specifications also list rate limits of 150 requests per second and 50 million tokens per minute, with API availability in and . Cursor's launch announcement similarly characterizes Grok 4.6 as designed for long-running agents and ambitious interactive and visual work, giving developers immediate access to the model inside an established coding-agent environment rather than requiring them to build a new harness around the API first. For enterprise buyers, that distribution may matter almost as much as another leaderboard result. Models increasingly compete not just on reasoning scores but on whether developers can place them inside existing coding, research and operational workflows without destabilizing those workflows or dramatically increasing inference costs, as well as incurring any blowback from associating with a controversial brand. Grok 4.6 does not establish an uncontested performance lead. Its launch instead presents a different proposition: frontier-level intelligence, large improvements over the previous generation, stronger long-running agent behavior and relatively aggressive token economics. The next test will be whether the efficiency Artificial Analysis observes on controlled agentic workloads carries into production. If Grok 4.6 can consistently complete long-running coding and knowledge-work tasks with fewer turns and fewer tokens, the model's most important benchmark may ultimately be the enterprise inference bill rather than the leaderboard.

SpaceXAI releases flagship Grok 4.6 model with advanced reasoning capabilities

SpaceXAI today released Grok 4.6, a large language model that it says can outperform Anthropic PBC's Claude Fable 5 in some areas. SpaceXAI was known as xAI until last month. The Elon Musk-founded artificial intelligence provider rebranded in connection with its acquisition by SpaceX Corp. In June, the combined company listed its shares on the Nasdaq via the biggest initial public offering on record. Grok 4.6 is rolling out only a month after SpaceXAI released its previous flagship LLM. According to the company, one of the main improvements is that its engineers spent more time training the former model. The extended training run used an AI-generated dataset designed to improve Grok 4.6's reasoning capabilities. SpaceXAI also provided the model with access to "high-quality engineering data." The initial training run was followed by two additional development steps. The first used a training method called supervised fine-tuning, or SFT, while the second used reinforcement learning. An SFT training run refines an LLM's output using a set of sample prompts and pre-packaged answers. Engineers mainly use the technique to ensure that LLM prompt responses are outputted in a user-friendly format. SpaceXAI used Grok 4.5, the predecessor of Grok 4.6, to optimize the latter model's SFT training phase. The optimization workflow focused on improving its ability to tackle science and programming tasks. SpaceXAI evaluated Grok 4.6 using the Artificial Analysis Intelligence Index. It's a dataset that combines nine popular AI benchmarks spanning fields such as science, coding and financial services. Grok 4.6 scored 61, which put it on par with OpenAI Group PBC's flagship GPT-5.6 Sol model and one point behind Claude Fable 5. SpaceXAI also compared the models across nine other benchmarks. Grok 4.5 managed to outperform Claude Fable 5 in three. One of the benchmarks, AA-Briefcase, evaluates LLMs' ability to perform knowledge work projects that would take a human weeks to complete. The other two evaluations comprised tasks spanning more than a half-dozen industries. SpaceXAI says that Grok 4.5 is particularly adept at generating software prototypes based on high-level descriptions. Additionally, it's better than its predecessor at creating visual assets such as interfaces. The model is also more likely to check its work for errors when working on long-horizon projects. The standard version of Grok 4.6 is priced at $2 per million input tokens and $6 per million output tokens. SpaceXAI also offers a faster edition that costs twice as much. The LLM is available through Cursor, the vibe coding platform that the company bought for $60 billion in June, and an internally developed programming tool called Grok Build.

SpaceXAI launches Grok 4.6 as Musk plans to train future models on employee data

SpaceXAI launched Grok 4.6 on Wednesday, and Elon Musk said future versions will be trained on "the sum total of all SpaceX information," including employee output and thinking. Grok 4.6 scored 61 on the Artificial Analysis Intelligence Index, matching OpenAI's GPT-5.6 Sol and trailing Anthropic's Claude Fable 5 by one point. SpaceXAI said the performance gains came from a longer supplemental training run and improved supervised fine-tuning and reinforcement learning, rather than a larger model. Grok 4.6 keeps the same 1.5 trillion parameter V9 base as Grok 4.5, which launched publicly on July 8, 2026. The company said Grok 4.6 used Grok 4.5 to regenerate and filter supervised fine-tuning training trajectories across reasoning, agent workflows, STEM, software engineering and knowledge work. That process removed problematic examples before training, avoiding a new pre-training cycle. Artificial Analysis said Grok 4.6 scored five points above Grok 4.5's 56 and 23 points above Grok 4.3's 38. The firm also gave Grok 4.6 a GDPval-AA v2 Elo rating of 1753 on agentic tasks, behind only Claude Opus 5 and roughly level with Claude Fable 5 and Qwen3.8 Max. Artificial Analysis said Grok 4.6 completed knowledge-work tasks in about 53 turns and 0.5 billion input tokens on average. Claude Opus 5 took 103 turns and 2.0 billion input tokens on the same measure. Pricing stayed unchanged from Grok 4.5 at $2 per million input tokens and $6 per million output tokens. Grok 4.6 is available through Cursor, Grok Build, the API, OpenRouter, Vercel and Cloudflare, and SpaceXAI said it would double usage quotas in Cursor and Grok Build for the first week. At an all-hands meeting posted publicly to X this week, Musk told employees, "So in a way, it will be trained on you." He added, "You will effectively be the parents of the AI. It will inherit your thoughts and ideas and beliefs, and I think that's a good thing." SpaceX did not say what employee data it would use, how it would collect that data, or whether workers could opt out. The company did not respond to media requests for comment. The source material cited Meta's Model Capability Initiative as the only recent example of an employer using worker computer activity for AI training. That program, launched in April 2026, logged keystrokes and screen content on most U.S. employees' work computers and was paused in June after private conversations, performance data and transcriptions became visible across the company. Meta Chief Technology Officer Andrew Bosworth told employees at launch that "there is no option to opt out of this on your work provided laptop." More than 1,500 employees later signed a petition against the program. SpaceX employees have weaker legal recourse than many other U.S. private-sector workers because the National Labor Relations Board dismissed its complaint against the company on Feb. 9, 2026, after a National Mediation Board opinion placed SpaceX under the Railway Labor Act. Musk also said at the all-hands meeting that SpaceX's AI revenue would "definitely" exceed all other company revenue in September. SpaceX's Q2 2026 results showed AI revenue of $2.56 billion, compared with about $5.25 billion from Starlink connectivity and space products combined. SpaceX CFO Bret Johnsen said on the Aug. 4 earnings call that the company had contracted $6.7 billion in new cloud services revenue "over a six-month period that begins ramping starting in October." Musk said on the same earnings call that Grok 4.7 is expected in three to four weeks with a target of 2.1 trillion parameters, and Grok 5 is planned before the end of 2026.

Elon Musk's SpaceXAI Enters The Big League With Grok 4.6, Offering Fable 5-Level Performance At An 80 Percent Discount

Elon Musk's SpaceXAI was largely written off as a has-been AI enterprise a while back. And yet, it has proven its detractors egregiously wrong with the all-new Grok 4.6, an AI model that goes toe-to-toe with Anthropic's Fable 5 and OpenAI's GPT-5.6 Sol at a fraction of price. Grok 4.6 achieves a score of 61 on Artificial Analysis Intelligence Index, which is the same as GPT-5.6 Sol and just a point shy of Claude Fable 5 SpaceXAI's Grok 4.6 offers comparable performance to Anthropic's Claude Fable 5 and OpenAI's GPT-5.6 Sol when it comes to Artificial Analysis Intelligence Index. It also tops the leadership board on GDPVal-AA v2, AA-Briefcase, and Harvey LAB (Vals) benchmarks. What's more, Grok 4.6 is around half the price of other frontier models at $2 per 1 million of input tokens and $6 per 1 million of output tokens. Just today, DeepSeek released its V4-Pro model, pricing it at $0.435 per 1 million tokens of input and $0.87 per 1 million tokens of output. Meanwhile, as we reported recently, SpaceX will bring around 2 gigawatts of compute online by the end of 2026, while targeting "closer to 10 than 5" gigawatts of compute by the end of 2027, replete with a tentative 20 gigawatt of power-and-cooling load. Do note that, as of May 2026, SpaceXAI's Colossus 1 data center supported over 220,000 NVIDIA GPUs, divided between H100, H200, and around 30,000 units of the GB200 AI accelerators. Meanwhile, the Colossus 2 data center currently boasts of over 550,000 GPUs, divided between GB200 and GB300 accelerators. Follow Wccftech on Google to get more of our news coverage in your feeds.

xAI introduces Grok 4.6 with long-running agents and interactive AI capabilities

xAI has introduced Grok 4.6, its latest AI model focused on long-running agents and interactive and visual work. The model builds on Grok 4.5 and is designed to handle complex tasks across multiple steps, including researching topics, analyzing information, working across codebases, and turning ideas into applications or other work artifacts. Grok 4.6 achieves frontier intelligence across several agentic coding and knowledge-work benchmarks. It matches GPT-5.6 Sol on the Artificial Analysis Intelligence Index, a composite score based on nine benchmarks. Training Grok 4.6 Grok 4.6 underwent a longer supplemental training run than Grok 4.5, using: * Curated model-generated data for reasoning and advanced technical concepts * High-quality engineering data * An improved optimizer and training recipe This provided the foundation for the supervised fine-tuning (SFT) and reinforcement learning (RL) stages that followed. xAI then used Grok 4.5 to regenerate SFT trajectories across different reasoning efforts, agent harnesses, and domains such as STEM, software engineering, and knowledge work. Problematic traces were filtered using model-based checks. The resulting SFT checkpoint showed strong performance and improved behavior. Grok 4.6 was trained on a wide range of agentic RL tasks, including: * Knowledge work * General coding * Kernel optimization * Web development * Computer-aided design (CAD) * More domain-specific environments Long-running and interactive projects xAI tested Grok 4.6 on projects designed to assess its range and ability to sustain work across many steps. The model can take a broad product idea and: * Research unfamiliar domains * Structure an application * Implement core interactions * Refine the result through several rounds of feedback On longer trajectories, xAI observed more self-testing and verification, with the model checking its work before moving on. Grok 4.6 also produced stronger first passes on visual and interactive projects than xAI typically saw with Grok 4.5. Given a concrete product idea, the model can establish an application's structure and visual language in one pass, then continue iterating on the result. Safety and safeguards xAI says Grok 4.6's safeguards have been improved and calibrated in line with the model's capabilities. The safety work covers legitimate use cases such as: * Vulnerability patching * Accelerating the engineering design cycle * Augmenting AI research xAI's safeguard evaluation included its widest-ever suite of pre-deployment testing for capabilities and safeguard calibration, along with extensive post-deployment and third-party testing. Pricing and availability Grok 4.6 is available today in Cursor and Grok Build. It is also available through the API and platforms including OpenRouter, Vercel, and Cloudflare. Pricing starts at: * Input: $2 per million tokens * Output: $6 per million tokens * Fast variant: 2× the standard price xAI is offering 2× included usage in Grok Build and Cursor for the first week.