Benchmarks
Deepswe
Models scored
37
Best score
0.73
Metric
Pass@1
Deepswe is an AI evaluation tracked by Epoch AI, with 37 scored model versions on record. The scores here are reported by third parties rather than produced by Epoch AI running the evaluation itself.
Scored models span February 2026 to July 2026. The best score recorded is 0.73, measured as Pass@1. Models come from OpenAI, Anthropic, Meta AI, Z.ai (Zhipu AI), Moonshot, Google DeepMind.
Full record
- Score metric
- Pass@1
- Models scored
- 37
- Best score recorded
- 0.73
- Earliest model scored
- Feb 17, 2026
- Most recent model scored
- Jul 9, 2026
- Third-party reported
- Yes
Leaderboard
| 01 | gpt-5.6-sol_max | OpenAI | 0.73 |
| 02 | gpt-5.6-sol_xhigh | OpenAI | 0.71 |
| 03 | claude-fable-5_xhigh | Anthropic | 0.7 |
| 04 | claude-fable-5_max | Anthropic | 0.7 |
| 05 | gpt-5.6-terra_max | OpenAI | 0.7 |
| 06 | gpt-5.6-sol_high | OpenAI | 0.69 |
| 07 | claude-fable-5_high | Anthropic | 0.69 |
| 08 | gpt-5.6-luna_max | OpenAI | 0.67 |
| 09 | gpt-5.5_xhigh | OpenAI | 0.67 |
| 10 | gpt-5.5_high | OpenAI | 0.64 |
| 11 | gpt-5.6-sol_medium | OpenAI | 0.61 |
| 12 | gpt-5.6-terra_xhigh | OpenAI | 0.6 |
| 13 | claude-opus-4-8_max | Anthropic | 0.59 |
| 14 | gpt-5.6-luna_xhigh | OpenAI | 0.57 |
| 15 | claude-opus-4-8_xhigh | Anthropic | 0.54 |
| 16 | gpt-5.5_medium | OpenAI | 0.54 |
| 17 | claude-sonnet-5_max | Anthropic | 0.54 |
| 18 | gpt-5.6-terra_high | OpenAI | 0.54 |
| 19 | muse-spark-1.1 | Meta AI | 0.53 |
| 20 | claude-opus-4-8_high | Anthropic | 0.52 |
| 21 | gpt-5.4-2026-03-05_xhigh | OpenAI | 0.52 |
| 22 | claude-sonnet-5_xhigh | Anthropic | 0.5 |
| 23 | claude-opus-4-8_medium | Anthropic | 0.49 |
| 24 | claude-sonnet-5_high | Anthropic | 0.48 |
| 25 | gpt-5.6-sol_low | OpenAI | 0.45 |
| 26 | gpt-5.6-luna_high | OpenAI | 0.44 |
| 27 | glm-5.2_max | Z.ai (Zhipu AI) | 0.44 |
| 28 | claude-opus-4-8_low | Anthropic | 0.41 |
| 29 | glm-5.2_high | Z.ai (Zhipu AI) | 0.36 |
| 30 | gpt-5.6-terra_medium | OpenAI | 0.35 |
| 31 | kimi-k2.7-code | Moonshot | 0.31 |
| 32 | claude-sonnet-4-6_high | Anthropic | 0.3 |
| 33 | gpt-5.5_low | OpenAI | 0.27 |
| 34 | gpt-5.6-terra_low | OpenAI | 0.24 |
| 35 | gemini-3.1-pro-preview | Google DeepMind | 0.12 |
| 36 | gpt-5.6-luna_medium | OpenAI | 0.11 |
| 37 | gpt-5.6-luna_low | OpenAI | 0.02 |
More benchmarks
SourceEpoch AI, 'AI Benchmarking Hub'. Published online at epoch.ai. Retrieved 2026-07-29 from https://epoch.ai/benchmarks. Licensed under CC BY 4.0.
← All benchmarks