xAI Releases Grok 4.7 Model, Focused on Coding and Knowledge Work

Tapbit news, September 21 — According to an xAI announcement, Grok 4.7 uses a larger base model and has undergone longer reinforcement learning training geared toward complex long-horizon tasks, strengthening its coding, long-context, and self-verification capabilities. The model is now available on Cursor, Grok Build, and the API, with input/output prices of $2 and $6 per million tokens respectively; xAI says it shows significant improvement over Grok 4.6 across multiple coding, office, and safety benchmarks.