xAI ships Grok 4.7 on a larger base model
xAI released Grok 4.7 on September 21, 2026, calling it its most capable model for coding and knowledge work. It sits on a new, larger base model with a longer reinforcement learning run weighted toward multi-hour tasks, and xAI reports 46.3% on CursorBench 4.0. Pricing stays at $2 per million input tokens and $6 output, the same as Grok 4.6, with a 500K context window. It is live in Cursor, Grok Build, the API and GitHub Copilot.
Why it matters
It scores 46 on the Intelligence Index at a $6 output price, which is well under what the Claude and GPT models above it charge. For long coding runs, output tokens are most of the bill.
Who should care
Teams running long coding agents who pay mostly for output tokens
What you can do
If you are on Grok 4.6, switch the model name: the price and speed are unchanged.