xAI has released Grok 4.6, the successor to Grok 4.5, with a particular focus on long-running agents and more ambitious interactive and visual work. The model is designed to stay with complex tasks across many steps, whether researching a topic, analyzing information, or producing work artifacts.
On benchmarks, 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 of nine benchmarks, landing at 61 points alongside GPT-5.6 Sol Max, just behind Fable 5 Max at 62 and ahead of Grok 4.5 High at 56.
For founders, the release signals that the agentic race is moving from single-shot chat to sustained multi-step work. Long-running agents need models that hold context, recover from errors, and keep producing across hours of tool use. Grok 4.6 is available now in the API console and across Grok apps, and xAI is positioning it as the model for serious autonomous work.