Ring-2.5-1T model card (Hugging Face) 10 Feb 2026
- Best of IMOAnswerBench (Mean@8) 86.69 / Heavy Thinking 90.00 90
- Best of AIME 26 (Mean@64) 93.75 / Heavy Thinking 95.00 95
- Best of HMMT (Nov. 25) (Mean@64) 97.34 / Heavy Thinking 98.33 98.3
- Best of LiveCodeBench-v6 (Mean@4) 87.89 / Heavy Thinking 89.40 89.4
- Gaia2-search (Mean@3) 75
- Tau2-Bench (Mean@3) 78.4
- SWE-Bench Verified (Accuracy) 72.4
- ARC-AGI-V2 (Pass@4) 19.3
- IMO 2025 (full score 42), self-tested 35
- CMO 2025 (full score 126), self-tested 105