OpenCompass Code - Code QA - English (CompassBench 2504): leaderboard
OpenCompass CompassBench 2504 edition (file dated 2025-05-13): code question answering sub-task of the code category, English questions. 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. 28 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Gemini 2.5 Pro (Preview 05-06) | 83.1 | #156 |
| 2 | Grok 3 | 77.7 | #313 |
| 3 | Gemma 3 27B (IT) | 76.5 | #528 |
| 4 | Qwen 3 235B A22B (Thinking) | 74.8 | #323 (Qwen 3 235B A22B) |
| 5 | Qwen 3 32B (Thinking) | 73.3 | #449 (Qwen 3 32B) |
| 6 | Qwen 3 235B A22B | 72.5 | #323 |
| 7 | Step 2 16K | 71.6 | #442 |
| 8 | Doubao-1.5-thinking-pro-250415 | 71.6 | #226 |
| 9 | MiniMax-Text-01 | 69.9 | #519 |
| 10 | DeepSeek V3 (0324) | 67.2 | #349 |
| 11 | GLM-Z1-32B-0414 | 65.7 | #405 |
| 12 | GPT-4.1 (2025-04-14) | 65.4 | #257 |
| 13 | GPT-4.1 Mini | 63.7 | #363 |
| 14 | ERNIE-X1-Turbo-32K | 62.3 | #288 |
| 15 | Spark X1 | 60.8 | #435 |
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-code-qa-english-compassbench-2504 · How It Works · Data refreshed daily, snapshot 2026-11-05.