CLEM ReferenceGame — leaderboard
Metric: Game Clemscore (%). Source: huggingface.co. 31 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Sonnet 4.5 | 100 |
| 2 | GLM-4.7 | 100 |
| 3 | Gemini 3 Flash | 100 |
| 4 | GPT-5.2 (High) | 100 |
| 5 | GPT-OSS-20B | 98.89 |
| 6 | GPT-5.2 (Medium) | 98.89 |
| 7 | Claude Sonnet 4.5 (High) | 98.89 |
| 8 | GPT-5.2 | 97.78 |
| 9 | Kimi K2 (Thinking) | 97.78 |
| 10 | GLM-4.6 | 94.45 |
| 11 | MiniMax-M2.5 | 86.67 |
| 12 | Qwen 3 Max | 86.67 |
| 13 | Mistral Large 3 | 84.44 |
| 14 | GLM-5 (Non-reasoning) | 76.67 |
| 15 | Qwen 2.5 Coder 32B Instruct | 73.33 |
Interactive version: theaggregate.ai/benchmark?slug=clem-referencegame · How the rankings work · Data refreshed daily, snapshot 2026-07-22.