Grok 4.6

SpaceXAI's frontier model, tuned for agents that run long multi-step jobs and for building interactive and visual work.

500k context window, text and image inputs, text-only output, with reasoning effort set to low, medium, high, or xhigh
API pricing is $2 per million input tokens and $6 per million output tokens below 200k prompt tokens, and double that above it; cached input is $0.50
Available in Cursor, Grok Build, the SpaceXAI API, and through OpenRouter, Vercel, and Cloudflare
SpaceXAI reports it level with the current frontier models on a composite of nine public benchmarks
On longer runs it tests and verifies its own work before moving on, which SpaceXAI reports as new in this version