OpenCompass Code - Generation (CompassBench 2411): leaderboard
OpenCompass CompassBench 2411 edition (file dated 2024-12-02): code generation sub-task of the code category. CompassBench is OpenCompass's periodic evaluation of chat models on a private question set; each edition is a separate board. Score 0-100, higher is better.
Metric: Score (%). Source: rank.opencompass.org.cn. Saturation forecast: Estimated already saturated. 34 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | O1 Mini (2024-09-12) | 56.4 | #412 |
| 2 | Claude 3.5 Sonnet (20241022) | 32.7 | #302 |
| 3 | GLM-4 Plus | 31.7 | #398 |
| 4 | GPT-4o (2024-11-20) | 30.7 | #392 |
| 5 | Mistral Large 2 (Nov) Instruct (2411) | 28.7 | #480 |
| 6 | Llama 3.1 405B Instruct FP8 | 27.7 | #400 |
| 7 | Qwen 2.5 72B Instruct | 26.7 | #469 |
| 8 | Grok Beta | 26.7 | #497 |
| 9 | 360gpt2-pro | 26.7 | #557 |
| 10 | DeepSeek V2.5 | 23.8 | #494 |
| 11 | TeleChat2 | 23.8 | #599 |
| 12 | GPT-4o (2024-08-06) | 22.8 | #343 |
| 13 | Yi Lightning | 20.8 | #429 |
| 14 | Step 2 16K | 18.8 | #442 |
| 15 | Gemma 2 27B (IT) | 17.8 | #652 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=opencompass-code-generation-compassbench-2411 · How It Works · Data refreshed daily, snapshot 2026-11-05.