Static score fallback. If the interactive dashboard does not appear, these tables confirm the deployed content is present.
Uncontaminated manual-evaluation scores from pre-release models and official human references for the core international competitions.
| Model | Overall /100 | IBO 2023 /453 | IChO 2025 /60 | IPhO 2025 /30 |
|---|---|---|---|---|
| Human No.1 | 93.26 | 395.4 | 57.1 | 29.2 |
| Human Gold | 72.43 | 354.7 | 36.6 | 23.4 |
| Human Silver | 56.4 | 293.1 | 28.7 | 17 |
| Human Bronze | 45.76 | 249.7 | 26.3 | 11.5 |
| Gemini 2.5 Pro | 77.06 | 404 | 37 | 24.1 |
| GPT-5 | 70.95 | 395 | 31.8 | 21.8 |
| GLM-4.5 | 58.76 | 329 | 19.8 | 21.2 |
| DeepSeek V3.1 | 59.29 | 312 | 26.6 | 19.4 |
| Qwen3-235B-A22B | 56.08 | 337 | 25.9 | 15.2 |
Latest multi-model results with source-linked problem details.
| Model | Overall /100 | IBO 2023 /453 | APhO 2025 Theory /30 | EuPhO 2025 Theory /30 | USNCO 2025 Part II /100 | IChO 2025 /60 | IPhO 2025 /30 | CPhO 2025 /320 | CChO 2025 /100 | INChO 2026 /106 | NBPhO 2026 /72 | USAPhO 2026 /150 |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Gemini 3 Flash 2025-12-17 | 82.69 | 428 | 26.75 | 23.6 | 93.5 | 44.74 | 25.2 | 300.5 | 59.55 | 81.5 | 53.15 | 136.5 |
| Doubao Seed 2.0 Pro 2026-02-14 | 67.97 | 415 | 21.72 | 13.6 | 72.3 | 37.14 | 18.92 | 243.7 | 53.18 | 87.2 | 42.33 | 106 |
| Gemini 3.1 Pro 2026-02-19 | 88.36 | 431 | 28.5 | 22.5 | 97 | 48.95 | 29.45 | 303.5 | 72.28 | 95.95 | 56.93 | 140 |
| GPT-5.4 2026-03-05 | 67.09 | 358 | 22.01 | 14.3 | 76.9 | 31.7 | 20.94 | 235.3 | 47.79 | 83.5 | 41.95 | 120 |
| GLM-5.1 2026-04-01 | 75.95 | 349 | 28.4 | 8.2 | 90 | 38.82 | 26.11 | 283 | 58.61 | 85.75 | 52.6 | 140.5 |
| Qwen3.6-Plus 2026-04-02 | 73.57 | 413 | 27.56 | 13.3 | 88.2 | 36.22 | 22.87 | 252.8 | 49.33 | 79.25 | 50.15 | 126.5 |
| MiniMax M2.7 2026-04-11 | 60.5 | 337 | 22.56 | 8.2 | 71.8 | 26.06 | 19.64 | 184.7 | 43.78 | 82.75 | 39.6 | 110 |
| Claude Opus 4.7 2026-04-16 | 75.23 | 407 | 24.19 | 16.5 | 75.2 | 44.49 | 22.29 | 262 | 56.59 | 82.7 | 48.9 | 141 |
| MiMo v2.5 Pro 2026-04-22 | 64.83 | 317 | 21.33 | 11.7 | 74 | 28.94 | 18.97 | 248.1 | 50.11 | 86 | 42.27 | 120.2 |
| GPT-5.5 2026-04-23 | 73.46 | 415 | 21.5 | 19 | 75.4 | 35.77 | 22.08 | 270.3 | 53.3 | 83 | 49 | 133 |
| DeepSeek V4 Flash 2026-04-24 | 67.86 | 324 | 20.44 | 20.4 | 94.5 | 38.33 | 20.53 | 182.9 | 46.46 | 83.8 | 43.9 | 102.5 |
| DeepSeek V4 Pro 2026-04-24 | 72.17 | 327 | 26.01 | 13.3 | 90 | 35.82 | 25.1 | 257.7 | 49.78 | 89.75 | 44.25 | 121.3 |
| Gemini 3.5 Flash 2026-05-19 | 85.62 | 430 | 27.91 | 16.8 | 95.7 | 47.98 | 26.41 | 307 | 70.6 | 92.25 | 61.42 | 143 |
| Qwen3.7-Max 2026-05-20 | 76.74 | 349 | 28.1 | 15.6 | 93 | 41.28 | 26.52 | 277.5 | 52.66 | 83 | 47.65 | 131 |