Grok 4.7 is xAI's most capable model for coding and knowledge work. It features a new larger base model trained with longer reinforcement learning on harder tasks, improving self-verification and longer context management. Grok 4.7 natively understands the Grok Bot harness, enhancing conversational tasks and general knowledge work.

On Cursor Bench 4.0, Grok 4.7 leads in price-performance and outperforms Grok 4.6, Fable 5.1, and GPT-5.6 Sol across software engineering, electrical engineering, multi-hour office and terminal work, legal, and clinical reasoning tasks.

Built with a new safeguard stack, Grok 4.7 demonstrates strongest refusals and jailbreak resistance. It leads on Latch Bio's biosafety benchmark at 62.4% and shows highest safety on Hacker Bench v0.3, allowing only 3.3% of risky dual-use prompts while rarely blocking legitimate security work.

Available today in Cursor, Grok Build, and via API, Grok 4.7 starts at $2 per million input tokens and $6 per million output tokens, with a fast variant at double price and speed.