Official benchmarks map / aime_2026
deepseek-ai/DeepSeek-V4.1-Flash and FINAL-Bench/Darwin-180B-RSI share first place on the aime_2026 leaderboard with 100 each, 0.8% ahead of zai-org/GLM-5.2. Read live from the Hugging Face API on November 8, 2026.
| # | Model | Organization | Score |
|---|---|---|---|
| 1 | deepseek-ai/DeepSeek-V4.1-Flash | deepseek-ai | 100 |
| 2 | FINAL-Bench/Darwin-180B-RSI | FINAL-Bench | 100 |
| 3 | zai-org/GLM-5.2 | zai-org | 99.2 |
| 4 | reflection/Beam-1.0-501B-A23B | reflection | 97.7 |
| 5 | thinkingmachines/Inkling | thinkingmachines | 97.1 |
| 6 | skt/A.X-K2 | skt | 97.1 |
| 7 | stepfun-ai/Step-3.5-Flash | stepfun-ai | 96.67 |
| 8 | moonshotai/Kimi-K2.6 | moonshotai | 96.4 |
| 9 | moonshotai/Kimi-K2.5 | moonshotai | 95.83 |
| 10 | zai-org/GLM-5 | zai-org | 95.83 |
aime_2026 is a dataset that Hugging Face marks as an official benchmark. Its leaderboard is built automatically from the .eval_results files that model repositories publish, so entries are mostly self-reported by model authors and evaluation settings differ between submissions. A verified tag appears only where Hugging Face marks an entry as verified.
Original board: huggingface.co/datasets/MathArena/aime_2026
Part of the Hugging Face official benchmarks map by QuantID. Unofficial. Not affiliated with or endorsed by Hugging Face.