| Rank | Model | Provider | Parameters | Context | Weights | |||||||||||||||||||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1 | Claude Opus 5.5 | Anthropic | Undisclosed | 1M | 2026-09-22 | Closed | — | — | — | — | — | 89.0 | — | 57.8 | — | — | — | — | 54.4 | 1846 | — | — | — | 67.7 | — | — | — | — | — | — | — | — | — | 81.8 | — | — | — | — | — | — | 66.4 | 58.7 | — | — |
| 2 | MiMo V2.6 Flash | Xiaomi | 310B MoE (15B active) | 1M | 2026-09-22 | Open | 27.6 | — | 52.3 | — | — | — | — | — | 95.1 | 67.9 | 25.3 | 6.0 | — | — | — | — | — | — | — | — | 61.2 | — | — | — | — | — | — | — | 80.8 | 26.0 | 47.5 | — | 87.6 | — | 28.8 | — | 73.6 | — |
| 3 | MiMo V2.6 Pro | Xiaomi | 1.02T MoE (42B active) | 1M | 2026-09-22 | Open | 31.6 | — | 53.1 | — | — | — | — | — | 94.0 | 71.9 | 47.9 | 17.8 | — | 1673 | — | — | — | — | — | — | 62.0 | — | — | — | — | — | — | — | 82.0 | 26.5 | 66.3 | — | 89.9 | — | 34.9 | — | 76.9 | — |
| 4 | DeepSeek V4.1 Flash | DeepSeek | 552B MoE (8B active) | 1M | 2026-09-10 | Open | 31.8 | — | 54.8 | 89.6 | — | 78.9 | 3471 | — | 88.1 | 74.2 | — | 15.3 | — | 1600 | 90.9 | — | 36.8 | 63.9 | — | — | 45.8 | — | — | 65.6 | — | — | 65.4 | — | — | 20.3 | 62.8 | — | 90.6 | 30.0 | 31.2 | — | — | 49.0 |
| 5 | GPT 6 Astra | OpenAI | Undisclosed | 1.05M | 2026-09-03 | Closed | 34.2 | — | 52.0 | — | — | — | — | — | — | 74.0 | 100.0 | 42.4 | 53.3 | 1542 | — | — | — | 57.2 | — | — | — | — | — | — | — | — | — | — | — | — | 85.4 | — | 89.9 | — | 57.9 | 64.6 | — | — |
| 6 | Claude Fable 5.1 | Anthropic | Undisclosed | 1M | 2026-09-01 | Closed | — | — | — | — | — | 88.4 | — | 51.8 | — | — | 83.0 | 30.4 | 50.3 | 1735 | — | — | — | 65.6 | — | — | — | — | — | — | — | — | — | 80.7 | — | — | — | — | 91.4 | — | 55.8 | 52.6 | 77.8 | — |
| 7 | GLM 5.3 Flash | Zhipu AI | 320B MoE (18B active) | 1M | 2026-08-26 | Open | — | — | — | — | — | — | — | — | — | 63.4 | — | — | — | — | — | — | — | 55.3 | — | — | — | — | — | — | — | — | 56.3 | — | — | — | — | — | 84.3 | — | — | — | — | — |
| 8 | Qwen3.8 Flash Next | Alibaba | 125B MoE (6B active) | 262K | 2026-08-26 | Open | — | — | — | — | — | — | — | — | — | 58.7 | — | — | — | — | 91.7 | — | 35.9 | — | — | — | 55.7 | 91.9 | — | — | — | — | 48.1 | — | — | — | — | 62.5 | — | — | — | — | — | — |
| 9 | GLM 5.3 | Zhipu AI | 753B MoE (40B active) | 1M | 2026-08-14 | Open | 28.5 | — | 48.8 | — | — | — | — | — | 84.5 | 66.9 | — | 15.0 | — | — | 88.1 | — | — | 62.5 | — | — | — | — | — | — | — | — | 58.0 | — | — | 19.0 | — | — | 88.2 | 28.3 | 37.9 | — | — | — |
| 10 | Qwen3.8 27B | Alibaba | 27B dense | 262K | 2026-08-14 | Open | — | — | — | — | — | — | — | — | — | 42.2 | — | — | — | — | 89.2 | — | 30.8 | — | — | — | 33.4 | 90.3 | — | — | — | — | 42.3 | — | 84.3 | — | — | 61.7 | 73.0 | — | — | — | — | — |
| 11 | DeepSeek V4 Pro 0813 | DeepSeek | 1.6T MoE (49B active) | 1M | 2026-08-13 | Open | 25.7 | — | 43.2 | — | — | — | 3348 | — | 83.3 | 62.7 | — | 5.4 | — | — | 92.4 | — | — | 60.0 | — | — | — | — | — | 65.3 | — | — | 61.5 | — | — | 15.5 | 56.4 | — | 87.9 | 11.8 | 12.4 | — | — | — |
| 12 | Qwen3.8 Max | Alibaba | 2.4T MoE (95B active) | 1M | 2026-08-02 | Closed | — | — | — | — | — | — | — | — | — | 56.6 | — | — | — | — | 92.6 | — | 43.6 | 56.2 | — | — | 53.4 | — | — | — | — | — | 55.9 | — | — | — | — | 67.7 | 86.6 | — | — | — | — | — |
| 13 | DeepSeek V4 Flash 0731 | DeepSeek | 284B MoE (13B active) | 1M | 2026-07-31 | Open | 25.2 | — | 37.7 | — | — | — | 3289 | — | 76.7 | 54.4 | — | 1.8 | — | — | 89.9 | — | — | 51.5 | — | — | — | — | — | 58.6 | — | — | 54.2 | — | — | — | 30.9 | — | 82.7 | 7.6 | 7.0 | — | — | — |
| 14 | Claude Opus 5 | Anthropic | Undisclosed | 1M | 2026-07-24 | Closed | 28.6 | — | 50.3 | 94.1 | — | 84.0 | — | 46.6 | — | 74.0 | 70.0 | 22.1 | 48.0 | 1708 | 93.4 | — | 56.3 | 63.6 | — | — | 65.7 | — | — | — | — | — | 75.3 | 74.0 | 83.4 | 37.0 | — | — | 89.1 | 43.3 | 51.8 | 29.0 | 80.6 | 52.0 |
| 15 | Kimi K3 | Moonshot AI | 2.8T MoE | 1M | 2026-07-16 | Open | 27.6 | — | 46.7 | 85.7 | — | 68.1 | — | — | 80.0 | 67.5 | 32.2 | 8.1 | — | 1524 | 92.9 | — | 43.5 | 59.8 | — | — | 54.3 | — | — | 65.6 | — | — | 58.0 | — | 84.8 | 17.5 | — | — | 88.3 | 17.7 | 12.6 | — | 76.5 | 41.0 |
| 16 | GPT 5.6 Sol | OpenAI | Undisclosed | 1.05M | 2026-07-09 | Closed | 26.7 | — | 45.8 | 88.9 | — | 79.9 | — | 41.7 | 84.5 | 73.0 | 78.5 | 33.7 | 47.5 | 1588 | 94.1 | — | 44.5 | — | — | — | 45.4 | — | — | — | — | — | 56.8 | — | 83.0 | 23.0 | 74.3 | — | 88.8 | 34.4 | 39.9 | 22.4 | 74.9 | 53.0 |
| 17 | Claude Fable 5 | Anthropic | Undisclosed | 1M | 2026-06-09 | Closed | 25.7 | — | 46.2 | — | — | — | — | — | — | 70.0 | 78.0 | 28.4 | — | 1595 | — | — | — | — | — | — | 57.4 | — | — | — | — | — | — | — | 86.0 | 33.0 | — | — | 84.3 | — | 42.4 | — | 77.9 | — |
| 18 | MiMo V2.5 Pro | Xiaomi | 1.02T MoE (42B active) | 1M | 2026-04-27 | Open | 13.2 | — | 16.0 | — | — | — | — | — | 40.0 | 19.0 | 16.6 | 0.1 | — | 1107 | — | — | — | — | — | — | 25.0 | — | — | — | — | — | — | — | — | 12.5 | 17.7 | — | 65.2 | — | 1.5 | — | 49.1 | — |
| 19 | Qwen3.6 27B | Alibaba | 27B dense | 262K | 2026-04-22 | Open | — | — | — | — | — | — | — | — | — | 13.3 | — | — | — | — | 87.8 | — | 24.0 | — | — | — | 21.8 | 83.9 | — | — | — | 86.2 | 36.2 | — | 63.9 | — | — | 53.5 | — | — | — | — | — | — |
| 20 | Qwen3.6 35B A3B | Alibaba | 35B MoE (3B active) | 262K | 2026-04-16 | Open | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 86.0 | — | 21.4 | — | — | — | — | 80.4 | — | — | — | 85.2 | 29.4 | — | — | — | — | 49.5 | — | — | — | — | — | — |
| 21 | Qwen3.6 Plus | Alibaba | Undisclosed | 1M | 2026-04-01 | Closed | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 90.4 | — | 28.8 | 50.6 | — | 94.3 | — | 87.1 | — | — | — | 88.5 | 37.9 | — | 62.5 | — | — | 56.6 | — | — | — | — | — | — |
| 22 | Claude 3.5 Sonnet | Anthropic | Undisclosed | 200K | 2024-10-22 | Closed | — | 1268 | — | — | 93.1 | — | — | — | — | — | — | — | — | — | 65.0 | 96.4 | — | — | 92.0 | 88.0 | — | — | 71.1 | — | 88.7 | 78.0 | — | — | — | — | — | — | — | — | — | — | — | — |
| 23 | Qwen2.5 72B | Alibaba | 72B | 128K | 2024-09-19 | Open | — | 1249 | — | — | 86.8 | — | — | — | — | — | — | — | — | — | 49.0 | 95.8 | — | — | 86.6 | 84.0 | — | — | 83.1 | — | 86.1 | 71.1 | — | — | — | — | — | — | — | — | — | — | — | — |
| 24 | o1 mini | OpenAI | Undisclosed | 128K | 2024-09-12 | Closed | — | 1230 | — | — | — | — | — | — | — | — | — | — | — | — | 60.0 | — | — | — | 92.4 | — | — | — | 90.0 | — | 85.2 | 74.0 | — | — | — | — | — | — | — | — | — | — | — | — |
| 25 | o1 preview | OpenAI | Undisclosed | 128K | 2024-09-12 | Closed | — | 1258 | — | — | — | — | — | — | — | — | — | — | — | — | 78.3 | — | — | — | 92.4 | — | — | — | 94.8 | — | 90.8 | 79.2 | — | — | — | — | — | — | — | — | — | — | — | — |
| 26 | DeepSeek V2.5 | DeepSeek | 236B MoE (21B active) | 128K | 2024-09-05 | Open | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 48.0 | 91.0 | — | — | 89.0 | 84.6 | — | — | 72.7 | — | 80.6 | 69.4 | — | — | — | — | — | — | — | — | — | — | — | — |
| 27 | GLM 4 Plus | Zhipu AI | Undisclosed | 128K | 2024-08-29 | Closed | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 81.0 | — | — | — | — | — | 86.0 | 68.0 | — | — | — | — | — | — | — | — | — | — | — | — |
| 28 | Mistral Large 2 | Mistral AI | 123B | 128K | 2024-07-24 | Open | — | 1224 | — | — | 78.9 | — | — | — | — | — | — | — | — | — | 48.0 | 93.0 | — | — | 92.0 | — | — | — | 68.6 | — | 84.0 | 61.2 | — | — | — | — | — | — | — | — | — | — | — | — |
| 29 | Llama 3.1 405B | Meta | 405B | 128K | 2024-07-23 | Open | — | 1225 | — | — | 93.0 | — | — | — | — | — | — | — | — | — | 51.1 | 96.8 | — | — | 89.0 | 88.6 | — | — | 73.8 | — | 87.3 | 73.3 | — | — | — | — | — | — | — | — | — | — | — | — |
| 30 | Llama 3.1 70B | Meta | 70B | 128K | 2024-07-23 | Open | — | 1180 | — | — | 81.6 | — | — | — | — | — | — | — | — | — | 41.7 | 95.1 | — | — | 80.5 | 87.5 | — | — | 67.9 | — | 83.6 | 66.4 | — | — | — | — | — | — | — | — | — | — | — | — |
| 31 | Llama 3.1 8B | Meta | 8B | 128K | 2024-07-23 | Open | — | 1140 | — | — | 62.3 | — | — | — | — | — | — | — | — | — | 30.4 | 84.5 | — | — | 72.6 | 80.0 | — | — | 51.9 | — | 69.4 | 48.3 | — | — | — | — | — | — | — | — | — | — | — | — |
| 32 | GPT 4o mini | OpenAI | Undisclosed | 128K | 2024-07-18 | Closed | — | 1207 | — | — | 75.0 | — | — | — | — | — | — | — | — | — | 40.2 | 87.0 | — | — | 87.2 | 80.5 | — | — | 70.2 | — | 82.0 | 58.5 | — | — | — | — | — | — | — | — | — | — | — | — |
| 33 | Gemini 1.5 Flash | Undisclosed | 1M | 2024-05-14 | Closed | — | 1201 | — | — | 75.0 | — | — | — | — | — | — | — | — | — | 38.6 | 86.2 | — | — | 74.3 | — | — | — | 54.9 | — | 78.9 | 67.3 | — | — | — | — | — | — | — | — | — | — | — | — | |
| 34 | Gemini 1.5 Pro | Undisclosed | 2M | 2024-05-14 | Closed | — | 1255 | — | — | 89.2 | — | — | — | — | — | — | — | — | — | 46.2 | 91.7 | — | — | 84.1 | — | — | — | 67.7 | — | 85.9 | 75.8 | — | — | — | — | — | — | — | — | — | — | — | — | |
| 35 | GPT 4o | OpenAI | Undisclosed | 128K | 2024-05-13 | Closed | — | 1287 | — | — | 83.1 | — | — | — | — | — | — | — | — | — | 53.6 | 95.8 | — | — | 90.2 | 84.0 | — | — | 76.6 | — | 88.7 | 72.6 | — | — | — | — | — | — | — | — | — | — | — | — |
| 36 | Yi Large | 01.AI | Undisclosed | 200K | 2024-05-13 | Closed | — | 1195 | — | — | — | — | — | — | — | — | — | — | — | — | — | 89.0 | — | — | 80.0 | — | — | — | — | — | 84.6 | — | — | — | — | — | — | — | — | — | — | — | — | — |
| 37 | Claude 3 Opus | Anthropic | Undisclosed | 200K | 2024-03-04 | Closed | — | 1240 | — | — | 93.7 | — | — | — | — | — | — | — | — | — | 50.4 | 95.0 | — | — | 84.9 | 85.0 | — | — | 60.1 | — | 86.8 | 68.0 | — | — | — | — | — | — | — | — | — | — | — | — |
Figures are taken from public model cards, technical reports and vendor announcements, and are approximate where sources disagree. Results from different benchmark versions or tool-use settings never share a column: Terminal Bench 2.1, 3.0 and DeepSWE v1.1 are distinct benchmarks, and HLE with tools is separate from HLE without tools. Bold values mark the best score in a column.