AI BriefingGrok-2 Beta Released
Key point
xAI has released Grok-2 and Grok-2 mini in beta, with significantly improved performance.
Details
xAI announced Grok-2 and its lightweight version Grok-2 mini, both showing significantly improved performance over the previous model, Grok-1.5. Grok-2 delivers state-of-the-art performance in chat, coding, and reasoning, and when tested under the name 'sus-column-r' on the LMSYS Chatbot Arena, it outperformed Claude 3.5 Sonnet and GPT-4-Turbo.
The Grok-2 series is currently available in beta on the 𝕏 platform, and will also be offered via enterprise API later this month.
Key benchmark achievements include the following:
- Achieved significant performance improvements over previous models on major academic benchmarks such as GPQA, MMLU, MMLU-Pro, and MATH.
- Also showed strength in visual reasoning, recording top-tier performance on MathVista (visual math reasoning) and DocVQA (document-based question answering).
Internally, instruction-following ability and factual accuracy were verified through the AI Tutor, and the interface within the 𝕏 platform was redesigned to improve user experience.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Step 5 Preview Released: 600B Parameter MoE Model Highlights Agentic Tasks and Financial Analysis PerformanceQwen3.5-based Small Decision Model 'Kev' Released with TypeSafe API Compatibility and Performance Metrics