SuperCLUE General (May 2025) - Code Generation: leaderboard
Metric: Score. Source: www.superclueai.com. 40 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | O4 Mini (High) | 91.52 |
| 2 | Gemini 2.5 Pro (Preview 05-06) | 91.32 |
| 3 | O3 | 88.76 |
| 4 | Doubao-1.5-thinking-pro-250415 | 87.97 |
| 5 | Qwen 3 32B (Thinking) | 87.18 |
| 6 | Claude Opus 4 (Thinking) | 86.98 |
| 7 | Gemini 2.5 Flash (Preview 05-20) | 86.79 |
| 8 | DeepSeek R1 | 86.59 |
| 9 | Gemini 2.5 Flash (Preview 04-17) | 86.59 |
| 10 | SenseNova V6 Reasoner | 86.19 |
| 11 | DeepSeek V3 (0324) | 84.81 |
| 12 | ERNIE-X1-Turbo-32K | 82.25 |
| 13 | DeepSeek R1 Distill Qwen 32B | 75.54 |
| 14 | Llama 4 Maverick Instruct | 75.15 |
| 15 | Kimi K1.5 | 73.37 |
Interactive version: theaggregate.ai/benchmark?slug=superclue-general-may-2025-code-generation · How It Works · Data refreshed daily, snapshot 2026-09-19.