Grok 4.7 is the most capable model from SpaceXSI, the lab formerly known as xAI, for coding and knowledge work. It was released on September 21, 2026. SpaceXSI says it uses a new, larger base model than Grok 4.6, was trained with a longer reinforcement learning run on harder tasks, and is better at checking its own work and managing long context.
Key facts
| Item | Grok 4.7 |
|---|---|
| Maker | SpaceXSI (formerly xAI) |
| Released | September 21, 2026 |
| API model ID | grok-4.7 |
| Context window | 500,000 tokens |
| Knowledge cutoff | May 2026 |
| API price per million tokens (prompts under 200K tokens) | $2 input, $0.50 cached input, $6 output |
| API price per million tokens (prompts of 200K tokens or more) | $4 input, $1 cached input, $12 output |
Sources: SpaceXSI's Grok 4.7 announcement and models and pricing docs, checked October 24, 2026.
Pricing details
Grok 4.7 is served at the same price as Grok 4.6: $2 per million input tokens and $6 per million output tokens for prompts under 200,000 tokens. Once a prompt reaches 200,000 tokens, the whole request is billed at the higher rate of $4 and $12. A fast variant offers twice the output speed at twice the price.
Where to use Grok 4.7
- API: through the Grok API, documented on our xAI API page. SpaceXSI's docs recommend Grok 4.7 for both code and chat. Search tools must be enabled for current events, since the model has no live knowledge on its own.
- Coding tools: in Cursor and Grok Build, plus other coding harnesses, model routers and cloud platforms.
- Grok apps: in Grok on paid plans of The X Subscription. Plan details are in our Grok subscription plans guide.
Safety
SpaceXSI says Grok 4.7 ships with a new safeguard stack and calls it its strongest model yet on refusals and jailbreak resistance. Select cybersecurity partners get invite only access to its red team capabilities for defense research.
How it compares
At $2 and $6, Grok 4.7 costs the same as GPT-6.1 Sol on input and less on output, but its context window is half the size. Against Claude Fable 5.1, SpaceXSI's own table shows Fable ahead on several coding benchmarks at five times the input price. For how Grok stacks up as an assistant, read ChatGPT vs Claude vs Gemini vs Grok or Is the xAI API Free?, and see the release timeline.