Grok 4.6
xAI's flagship model for long-running coding agents and interactive visual work
xAI released Grok 4.6 on August 12, 2026, extending Grok 4.5’s coding focus toward longer-running agent sessions and more demanding interactive and visual tasks. The model carries a 500,000-token context window, accepts text and image input, and exposes configurable reasoning effort levels from low to xhigh so developers can trade off latency against depth on a given request. xAI priced it at $2 per million input tokens and $6 per million output tokens for prompts under 200,000 tokens, with rates roughly doubling once a request crosses that threshold.
The model reached the xAI API, the Grok app, and X at launch, and xAI added it to Microsoft Foundry shortly after, putting it alongside other frontier models available to enterprise customers through Microsoft’s platform. xAI is positioning Grok 4.6 mainly against coding-oriented models like Claude Opus 5 and GPT-5.6 Sol, pointing to strength on tool-calling and low hallucination rates in agentic settings rather than pure benchmark scores.