The official model card for Grok 4.5. It documents the model's capability benchmarks and safety evaluations.
Blog / researchJul 14, 2026research1 min readWe're publishing the official model card for Grok 4.5.
It documents the model's capability benchmarks and safety evaluations.
Related postsJul 8, 2026ResearchIntroducing Grok 4.5Cursor Team3 min readJul 20, 2026ResearchAgent swarms and the new model economicsWilson Lin17 min readApr 6, 2026ResearchBetter MoE model inference with warp decodeLess, Federico & Zhiyuan10 min readView more posts →
Grok 4.5 is SpaceXAI's smartest model built for coding, agentic tasks, and knowledge work.
Our most intelligent model and the first we've built for more than software engineering.
SpaceX's latest AI model, Grok 4.5, offers modern coding capabilities and agentic workflows at competitive pricing.
SpaceXAI launched Grok 4.5, a new model for coding, agentic tasks, and knowledge work, now available in Grok Build, Cursor, and…
Built for long-running agents and more ambitious interactive and visual work.
SpaceXAI released Grok 4.5, a Cursor-trained model for coding, agentic tasks, and knowledge work at $2 input pricing.
