| Rank | Model | Provider | Score |
|---|---|---|---|
| 1 | GPT 4o | OpenAI | 1287 |
| 2 | Claude 3.5 Sonnet | Anthropic | 1268 |
| 3 | o1 preview | OpenAI | 1258 |
| 4 | Gemini 1.5 Pro | 1255 | |
| 5 | Qwen2.5 72B | Alibaba | 1249 |
| 6 | Claude 3 Opus | Anthropic | 1240 |
| 7 | o1 mini | OpenAI | 1230 |
| 8 | Llama 3.1 405B | Meta | 1225 |
| 9 | Mistral Large 2 | Mistral AI | 1224 |
| 10 | GPT 4o mini | OpenAI | 1207 |
| 11 | Gemini 1.5 Flash | 1201 | |
| 12 | Yi Large | 01.AI | 1195 |
| 13 | Llama 3.1 70B | Meta | 1180 |
| 14 | Llama 3.1 8B | Meta | 1140 |
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. Only models with a published result are listed; the score column is shaded relative to the best and worst result on this benchmark.