Benchmarks
GSO
Models scored
36
Best score
0.44
Metric
Score OPT@1
GSO is an AI evaluation tracked by Epoch AI, with 36 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 October 2024 to April 2026. The best score recorded is 0.44, measured as Score OPT@1. Models come from Anthropic, OpenAI, Google DeepMind, Moonshot, Alibaba, Z.ai (Zhipu AI),Tsinghua University.
Full record
- Score metric
- Score OPT@1
- Models scored
- 36
- Best score recorded
- 0.44
- Earliest model scored
- Oct 22, 2024
- Most recent model scored
- Apr 23, 2026
- Third-party reported
- Yes
Leaderboard
| 01 | claude-opus-4-7 | Anthropic | 0.44 |
| 02 | claude-opus-4-7_high | Anthropic | 0.44 |
| 03 | claude-opus-4-6_high | Anthropic | 0.41 |
| 04 | gpt-5.5_xhigh | OpenAI | 0.4 |
| 05 | claude-opus-4-6 | Anthropic | 0.33 |
| 06 | claude-opus-4-6_unknown | Anthropic | 0.33 |
| 07 | gpt-5.4-2026-03-05_xhigh | OpenAI | 0.31 |
| 08 | gpt-5.2-2025-12-11_high | OpenAI | 0.27 |
| 09 | claude-opus-4-5-20251101 | Anthropic | 0.27 |
| 10 | claude-opus-4-5-20251101_unknown | Anthropic | 0.26 |
| 11 | gpt-5.4-2026-03-05_high | OpenAI | 0.25 |
| 12 | gemini-3.1-pro-preview | Google DeepMind | 0.23 |
| 13 | gemini-3-pro-preview | Google DeepMind | 0.19 |
| 14 | claude-sonnet-4-5-20250929_unknown | Anthropic | 0.15 |
| 15 | claude-sonnet-4-5-20250929 | Anthropic | 0.15 |
| 16 | gpt-5.1-2025-11-13_unknown | OpenAI | 0.14 |
| 17 | gpt-5.1-2025-11-13_high | OpenAI | 0.14 |
| 18 | gemini-3-flash-preview | Google DeepMind | 0.1 |
| 19 | o3-2025-04-16_high | OpenAI | 0.09 |
| 20 | claude-opus-4-20250514 | Anthropic | 0.07 |
| 21 | gpt-5-2025-08-07_high | OpenAI | 0.07 |
| 22 | claude-opus-4-20250514_unknown | Anthropic | 0.07 |
| 23 | Kimi-K2-Instruct | Moonshot | 0.05 |
| 24 | Qwen3-Coder-480B-A35B-Instruct | Alibaba | 0.05 |
| 25 | claude-sonnet-4-20250514_unknown | Anthropic | 0.05 |
| 26 | claude-sonnet-4-20250514 | Anthropic | 0.05 |
| 27 | claude-3-5-sonnet-20241022 | Anthropic | 0.05 |
| 28 | gemini-2.5-pro-preview-06-05 | Google DeepMind | 0.04 |
| 29 | gemini-2.5-pro | Google DeepMind | 0.04 |
| 30 | claude-3-7-sonnet-20250219_unknown | Anthropic | 0.04 |
| 31 | claude-3-7-sonnet-20250219 | Anthropic | 0.04 |
| 32 | o4-mini-2025-04-16_high | OpenAI | 0.04 |
| 33 | GLM-4.5-Air | Z.ai (Zhipu AI),Tsinghua University | 0.03 |
| 34 | o3-mini-2025-01-31_low | OpenAI | 0.01 |
| 35 | o3-mini-2025-01-31_high | OpenAI | 0.01 |
| 36 | gpt-4o-2024-11-20 | OpenAI | 0 |
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