Benchmark
AIME 2026
Math benchmarks measure how well models solve maths problems. Scores are percentages. Higher is better.
- Models measured
- 40
- Scores recorded
- 40
- Sources
- 3
- IQ weight
- 2.58
Score distribution
- 90.9%
- Mean
- 94.2%
- Median
- 11.2%
- Std dev
- 82.5%
- 10th percentile
- 97.1%
- 90th percentile
- 37.5%
- Min
This benchmark has a weight of 2.58 for chat models, measured over 40 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.
Leaderboard
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V4.1 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 100% |
| 2 | GLM 5.2· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 99.2% |
| 3 | Beam 1.0 501B A23B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 97.7% |
| 4 | Inkling FP4· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 97.1% |
| 5 | Sakana Namazu· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 96.7% |
| 6 | Step 3.5 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 96.7% |
| 7 | Kimi K2.6· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 96.4% |
| 8 | GLM 5· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 95.8% |
| 9 | Kimi K2.5· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 95.8% |
| 10 | Ring 2.6 1T· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 95.8% |
| 11 | Inkling Small· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 95.5% |
| 12 | GLM 5.1· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 95.3% |
| 13 | Qwen3.6 Plus· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 95.3% |
| 14 | Solar Pro 4· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 95.3% |
| 15 | Muse Glimmer 30B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 94.7% |
| 16 | Mai Thinking 1· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 94.5% |
| 17 | Doubao Seed 2.0 Pro· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 94.2% |
| 18 | MiniMax M2.7· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 94.2% |
| 19 | DeepSeek V3.2· OpenEvals Leaderboard data· Reported by the model's developer. The source relaying it did not verify it. | 94.2% |
| 20 | Qwen3.6 27B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 94.1% |
| 21 | Qwen3.5 397B A17B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 3 sources
| 93.3% |
| 22 | openPangu 2.0 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 93.3% |
| 23 | Qwen3 Max· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 93.3% |
| 24 | Ling 3.0 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 93.2% |
| 25 | Qwen3.6 35B A3B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claimValues from 2 sources
| 92.7% |
| 26 | Qwen3.5 27B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 90.8% |
| 27 | Nemotron 3 Super 120B A12B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 90% |
| 28 | Gemma 4 31B IT· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 89.2% |
| 29 | Gemma 4 26B A4B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.3% |
| 30 | Seed 2.0 Lite· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 88.3% |
| 31 | Qwen3 30B A3B2507, thinking· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 87.5% |
| 32 | ling-2.6-1t· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 87.4% |
| 33 | Seed 2.0 Mini· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 86.7% |
| 34 | MiniCPM5-2B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 86.5% |
| 35 | Qwen3-4B-Thinking-2507· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 82.5% |
| 36 | Gemma 4 12B It· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 77.5% |
| 37 | MiniCPM-V 4.7 35B-A3B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 76.5% |
| 38 | Ling 2.6 Flash· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim | 73.8% |
| 39 | Gemma 4 E4B-it· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 42.5% |
| 40 | Gemma 4 E2B-it· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim | 37.5% |
Data credits: Hugging Face official benchmark leaderboards; licence varies by benchmark · Data by LLM Stats · OpenEvals/leaderboard-data, MIT (licence)