Introducing Grok 4.6
x.ai released Grok 4.6, focusing on long-running agents and ambitious interactive and visual work. The model achieves frontier performance comparable to GPT-5.6 Sol on the Artificial Analysis Intelligence Index and is available in Cursor and Grok Build.
It's remarkable how quickly Elon Musk's AI initiative has caught up with major companies, positioning Grok 4.6 among the top models in the field alongside competitors like Mythos5 and GPT-5.6 Sol.
What are the key improvements in Grok 4.6 compared to the previous version?
Grok 4.6 focuses on long-running agents and ambitious interactive projects with visual capabilities. The model better handles multi-step tasks including research, analysis, codebase work, and transforming ideas into finished products. It also shows improved self-testing and verification during longer task sequences.
How does Grok 4.6 compare to competing models?
Grok 4.6 achieves comparable performance to GPT-5.6 Sol on the Artificial Analysis Intelligence Index. It matches GPT-5.6 Sol across agentic coding and knowledge work benchmarks and outperforms Fable 5 Max on metrics like CursorBench and FrontierCode.
Where is Grok 4.6 available and what is the pricing?
The model is available in Cursor and Grok Build, via API, and partner platforms including OpenRouter, Vercel, and Cloudflare. Pricing starts at $2 per million input tokens and $6 per million output tokens, with higher pricing for the faster variant.
How was Grok 4.6 trained?
The model underwent extended supplemental training beyond Grok 4.5 using high-quality data for reasoning and technical concepts, engineering data, and an improved optimizer. It followed with SFT stages using trajectories regenerated by Grok 4.5 and RL training on agentic tasks across domains like STEM and web development.
- Grok Bot: AI agents working 24/7 inside your existing tools — x.ai 76 % match
- GLM-5.3 from Z.ai leads in vulnerability detection and will soon be available for download — z.ai 74 % match
- LFM2.5-2.6B: small and capable local AI model — liquid.ai 73 % match
- Grok 4.6
- x.ai
- GPT-5.6 Sol8
- Grok 4.5
- Cursor4
- Fable 5 Max
- Artificial Analysis Intelligence Index
- Elon Musk5