Grok 4.7
By Steven Van ·
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.
Grok 4.7 is now available in the Grok API as grok-4.7, in Cursor, and in Grok Build, according to xAI's announcement. xAI says it's built for coding, agentic tasks, and knowledge work, and is served at the same price and speed as Grok 4.6.
- A 500k context window, with text and image input and text-only output, and no text output limit.
- Below 200k prompt tokens, pricing is $2 per million input tokens, $0.50 per million cached input tokens, and $6 per million output tokens.
- Reasoning effort can be set to low, medium, high (the default), or xhigh.
- Also served on xAI's US regional endpoint.
xAI says the model uses a larger base than Grok 4.6 and went through a longer reinforcement learning run weighted toward tasks that take many hours to complete, making it better at checking its own work and handling longer context. It was also trained to natively understand the Grok Bot harness. A faster variant, Grok 4.7 Fast, runs at twice the output speed for twice the price, but is available only through Cursor and Grok Build, not the public API.
xAI also says Grok 4.7 ships with a new safeguard stack it calls its strongest yet on refusals and jailbreak resistance, and that select cybersecurity partners now have invite-only access to its red-team capabilities for defense research.