SpaceXAI Launches Grok 4.6, Matching GPT-5.6 Sol on AI Index

SpaceXAI has released Grok 4.6, an updated AI model the company says delivers performance comparable to OpenAI's GPT-5.6 Sol and Anthropic's Claude Fable 5. The release focuses on long-running agents, coding, knowledge work, and more ambitious interactive and visual projects.



According to the official announcement, Grok 4.6 scored 61 on the Artificial Analysis Intelligence Index, matching GPT-5.6 Sol Max while finishing one point behind Claude Fable 5 Max. SpaceXAI also reported gains over Grok 4.5 across every evaluation it published, including CursorBench, FrontierCode, APEX-Agents, and Terminal-Bench.


SpaceXAI attributes the improvements to a longer supplemental training run, curated model-generated data for reasoning and advanced technical concepts, higher-quality engineering data, an improved optimizer and training recipe, and expanded reinforcement learning for coding and knowledge work. Grok 4.6 is better at turning broad ideas into working applications, researching unfamiliar domains, structuring projects, and refining results over multiple iterations, while increasingly checking its own work before moving forward.



Grok 4.6 is available now through Cursor, following SpaceXAI's move to acquire Cursor in June, as well as in Grok Build, the SpaceXAI API, OpenRouter, Vercel, and Cloudflare. API pricing starts at $2 per million input tokens and $6 per million output tokens, while a faster variant costs twice as much. Cursor and Grok Build users will receive double the included Grok 4.6 usage during the first week.