xAI ships Grok 4.6 with a 500,000-token context window
The new flagship model is built for long-running agentic work, coding, and knowledge tasks, and now matches OpenAI's top model.
Illustration · AI-generated (AI IN LIFE)
At a glance
- New model: Grok 4.6, succeeding Grok 4.5 after about one month
- Context window: up to 500,000 tokens
- Knowledge cutoff: February 1, 2026
- Intelligence index (High): 61 points, up from 56 for Grok 4.5
- Pricing: $2/1M input tokens, $6/1M output tokens, $0.50 for cached input
xAI has released Grok 4.6, its new flagship model, succeeding Grok 4.5 just about a month after that model's launch. It is available through the xAI API and major model gateways, and processes text and image inputs with text-only output and no fixed output limit.
Its headline feature is a context window of up to 500,000 tokens, built for long-running agentic work, coding tasks, and complex knowledge work. The model carries a knowledge cutoff of February 1, 2026.
On xAI's own intelligence index, Grok 4.6 in its High configuration scores 61 points, up from 56 for Grok 4.5, putting it level with OpenAI's GPT 5.6 Sol Max. Pricing is set at $2 per million input tokens and $6 per million output tokens, with a discounted $0.50 rate for cached input.
FAQ
What does a 500,000-token context window mean?
The model can process very large amounts of text in a single request, such as entire codebases or long document collections, without losing track of context.
Where is Grok 4.6 available?
Through xAI's own API and model gateways such as OpenRouter, Vercel, and Cloudflare.
How does Grok 4.6 compare to rivals?
It scores 61 on xAI's intelligence index, putting it level with OpenAI's GPT 5.6 Sol Max, OpenAI's previous top model.
