Grok 4.7

What is Grok 4.7?

Grok 4.7 is xAI’s flagship model for coding, agentic tasks and knowledge work, released on 21 September 2026. xAI’s announcement, published under the SpaceXSI name, says it uses a new, larger base model than Grok 4.6 and is served at the same price and speed. xAI’s documentation lists text and image input, text output, a 500,000-token context window, no text output limit, a May 2026 knowledge cutoff and four reasoning efforts: low, medium, high (the default) and xhigh.

Prompts under 200K tokens cost $2 per million input tokens, $0.50 for cached input and $6 per million output tokens. Once a prompt reaches 200K tokens, every token in that request is billed at $4, $1 and $12. The Batch API takes 20% off those rates, and the US regional endpoint adds 10% ($2.20 input, $6.60 output). Grok 4.7 Fast, the same model on faster infrastructure, costs twice the standard rates for prompts under 200K tokens, but it runs only in Cursor and Grok Build and is not offered on the public API (see xAI’s pricing page).

Developers call it as grok-4.7 on the xAI API. Microsoft Foundry and Amazon Bedrock list it, and xAI names OpenRouter, Vercel and Cloudflare as gateways; it is also the default model in Grok Build and available in Cursor. xAI’s earlier flagship, Grok 4, left the API on 15 May 2026, and calls to it now run on Grok 4.3, which costs $1.25 / $2.50 with a 1M-token window and is the cheaper choice when price matters more than capability. Our Grok API Pricing Guide covers the whole lineup, and the KI-API-Kostenrechner turns these rates into a monthly bill.