SpaceXAI has announced the release of Grok 4.6, the latest iteration of its generative AI chatbot. This version is designed to enhance performance in long-running agents, coding tasks, knowledge work, and complex visual projects.
According to the company, Grok 4.6 demonstrates intelligence on par with OpenAI’s GPT-5.6 Sol and Anthropic’s Claude Fable 5. On the Artificial Analysis Intelligence Index, Grok 4.6 achieved a score of 61, matching GPT-5.6 Sol’s maximum reasoning level and trailing Fable 5 Max by a single point.
Performance evaluations indicate that Grok 4.6 surpasses its predecessor, Grok 4.5, across various benchmarks, including CursorBench, FrontierCode, APEX-Agents, and Terminal-Bench. These improvements are attributed to an extended supplemental training period, enhanced engineering data, and expanded reinforcement learning focused on coding and knowledge work.
Grok 4.6 is now accessible through multiple platforms, including Cursor, Grok Build, the SpaceXAI API, OpenRouter, Vercel, and Cloudflare. API pricing is set at $2 per million input tokens and $6 per million output tokens, with a faster version available at double the cost. Additionally, Cursor and Grok Build users will receive double the included Grok 4.6 usage during its first week.
SpaceXAI’s recent acquisition of Cursor in June has further bolstered its AI capabilities. This week, the company also introduced Grok Bot, expanding its AI offerings to iPhone, Mac, and other platforms.
By releasing Grok 4.6, SpaceXAI continues to advance its position in the competitive AI landscape, offering a high-performing model at a competitive price point. This development underscores the rapid evolution of AI technologies and their increasing accessibility across various applications.
