Grok 4.7: Pricing, Context Window and Model Guide

What Grok 4.7 is

SpaceXSI released Grok 4.7 on September 21, 2026, as its most capable model for coding and knowledge work. It extends Grok 4.6, and the company says it completes longer and harder tasks on its own than any of its previous models, reaching results in fewer steps and with fewer output tokens than other frontier models. The developer documentation describes it as a flagship model for code and everything else, with agentic tool calling and configurable reasoning.

Pricing

Grok 4.7 costs $2 per million input tokens and $6 per million output tokens, the same as Grok 4.6. Several frontier models in its class charge $10 or more per million output tokens.

The Fast variant, and where you can get it

Grok 4.7 Fast is the same model served on faster infrastructure, at twice the standard token rates, or one and a half times for long-context requests. The Fast variant is only offered through Cursor and Grok Build; it is not available on the public xAI API. The standard model is available through the Grok API, Cursor, Grok Build, third-party coding tools, model routers and cloud platforms.

Context, output and reasoning

The context window is 500,000 tokens, half that of the GPT-6 and Claude 5.5 families. The documentation states there is no text output limit. The knowledge cutoff is May 2026. Reasoning can be set to low, medium, high or xhigh, with high as the default. Input is text and images, output is text, and a US regional endpoint is available.

Tools

Grok 4.7 supports function calling, web search, X search and code execution. X search lets the model search posts on X.

How it measures up

SpaceXSI's launch post reports 71.0% on DeepSWE v1.1, 46.3% on CursorBench 4.0 and 64.0% on EEBench, among others; those are the vendor's own runs. On LMArena's independent text board Grok 4.7 has 1443 Elo (rank 92, xhigh effort). Current figures with their confidence ranges are on our benchmarks page and update weekly.

Official Resources