xAI · Model release
Grok 4.7: a larger base model at Grok 4.6's price
Grok 4.7 is xAI's flagship since 2026-09-21 — a larger new base model with longer reinforcement learning on multi-hour tasks, sold at Grok 4.6's $2/$6 rate card.
What changed
- Priced $2/$6 per MTok ($4/$12 for every token once a prompt reaches 200K); cached input $0.50, or $1.00 at 200K and above. Context stays at 500K with no published output cap.
- The standard `grok-4.7` model is listed in xAI's public API, Cursor and Grok Build. GitHub announced a gradual Copilot rollout for eligible tiers, subject to admin policy; do not assume same-day access for every user.
- xAI's launch comparison reports 46.3% on CursorBench 4.0 at Extra High; this vendor-published, harness-specific result is not a general API ranking. No Artificial Analysis aggregate score is asserted here.
Best fit
- Coding agents in Cursor and eligible Copilot rollouts
- xAI API stacks
- Frontier work at the published token rate
What to be careful about
- 500K context is half the 1M-class frontier rows, and there is no Batch API.
- The xAI docs list a 1.1x US regional premium. Grok 4.7 Fast is a separate variant available only in Cursor and Grok Build, at 2x standard rates (1.5x for long-context requests).
Dataset snapshot: Grok 4.7 is also represented in the checked comparison dataset. Prices and benchmark figures carry their own field dates.