Grok model family (Grok 4 / 4.1 / 4.5 / 4.6 / 4.7)
Overview
xAI's frontier LLM line. Grok 4.7 is SpaceXAI's most capable model for coding and knowledge work, using a larger base than Grok 4.6 with a longer reinforcement-learning run on multi-hour tasks; Grok 4.6 (August 12, 2026) remains the prior flagship; Grok 4.1 Fast offers a 2M-token context window for agentic tool use.
Grok 4.7 launched in Cursor, Grok Build, the Grok API, third-party coding harnesses, and cloud platforms at $2 and $6 per million input and output tokens, with a faster variant at twice the output speed and twice the price. SpaceXAI said it ships a new safeguard stack (62.4% on LatchBio biosafety; 3.3% of risky HackerBench v0.3 prompts allowed through). Grok 4.6, released August 12, 2026, scored 61 on the Artificial Analysis Intelligence Index. Around September 13-14 Musk also named a 2.5T Grok 4.8 finishing pretraining then starting RL, a claim that has not appeared on x.ai. The roughly 6T-parameter Grok 5 remains in training on Colossus 2 with no confirmed release date.