Benchmarks
CL Bench
Models scored
23
Best score
0.28
Metric
Overall
CL Bench is an AI evaluation tracked by Epoch AI, with 23 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 April 2025 to March 2026. The best score recorded is 0.28, measured as Overall. Models come from OpenAI, xAI, Anthropic, Google DeepMind, Alibaba, Moonshot, Z.ai (Zhipu AI), Xiaomi Corp and others.
Full record
- Score metric
- Overall
- Models scored
- 23
- Best score recorded
- 0.28
- Earliest model scored
- Apr 16, 2025
- Most recent model scored
- Mar 31, 2026
- Third-party reported
- Yes
Leaderboard
| 01 | gpt-5.4-2026-03-05_xhigh | OpenAI | 0.28 |
| 02 | gpt-5.1-2025-11-13_high | OpenAI | 0.24 |
| 03 | grok-4-20 | xAI | 0.22 |
| 04 | gpt-5.1-2025-11-13_unknown | OpenAI | 0.21 |
| 05 | claude-opus-4-5-20251101_unknown | Anthropic | 0.21 |
| 06 | gemini-3.1-pro-preview | Google DeepMind | 0.21 |
| 07 | claude-opus-4-6_unknown | Anthropic | 0.21 |
| 08 | qwen3.6-plus | Alibaba | 0.2 |
| 09 | qwen3.5-plus | Alibaba | 0.2 |
| 10 | kimi-k2.5 | Moonshot | 0.19 |
| 11 | glm-5 | Z.ai (Zhipu AI) | 0.19 |
| 12 | gpt-5.2-2025-12-11_unknown | OpenAI | 0.18 |
| 13 | gpt-5.2-2025-12-11_high | OpenAI | 0.18 |
| 14 | o3-2025-04-16_high | OpenAI | 0.18 |
| 15 | moonshotai/Kimi-K2-Thinking | Moonshot | 0.18 |
| 16 | zai-org/glm-4.7 | Z.ai (Zhipu AI) | 0.16 |
| 17 | gemini-3-pro-preview | Google DeepMind | 0.16 |
| 18 | mimo-v2-pro | Xiaomi Corp | 0.16 |
| 19 | qwen3-max-2025-09-23 | Alibaba | 0.14 |
| 20 | DeepSeek-V3.2-Exp_thinking | DeepSeek | 0.13 |
| 21 | deepseek/deepseek-v3.2 | DeepSeek | 0.12 |
| 22 | kimi-k2-thinking | Moonshot | 0.12 |
| 23 | MiniMax-M2.5 | MiniMax | 0.11 |
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