xAI
2 stories · first 22 Sep · latest 5 Oct
Products and models
Facts we checked
Each fact was checked against the page it links to.
-
Grok 4.7, SpaceXAI's frontier model for coding, agentic tasks, and knowledge work, is now available on the xAI API as `grok-4.7`.
Release Notes · xAI · archived copy · checked 22 Sep 2026 · Cite this fact · the story
-
Grok 4.7 takes text and image inputs with text-only output
Release Notes · xAI · archived copy · checked 22 Sep 2026 · Cite this fact · the story
-
Above 200k prompt tokens, Grok 4.7 costs $4 / $1 / $12 per 1M tokens (input / cached input / output)
Release Notes · xAI · archived copy · checked 22 Sep 2026 · Cite this fact · the story
-
Grok 4.7 costs $2 / $0.50 / $6 per 1M tokens (input / cached input / output) below 200k prompt tokens
Release Notes · xAI · archived copy · checked 22 Sep 2026 · Cite this fact · the story
-
Grok 4.7 has a 500k context window
Release Notes · xAI · archived copy · checked 22 Sep 2026 · Cite this fact · the story
Numbers over time
Prices, context windows and test scores we checked. Each number links to the fact it comes from.
Grok 4.7: Benchmark score, CheatBench cheating rate with Grok Build (lower is better) (third party)
| Date | Value | Fact |
|---|---|---|
| 2026-10-05 | 78% | Cite this fact |
All 2 stories
Cases
Ongoing case: Frontier model API pricesOngoing case: Million-token context windowsOngoing case: The White House and AI rules
Its own pages
Pages the company published that our stories cite.
- Release Notes · docs.x.ai
Built from our published stories and the facts we checked, and kept up to date by them. How we correct.