SuperCLUE General (March 2026) - Code Generation: leaderboard
Metric: Score. Source: www.superclueai.com. 24 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Max) | 71.15 |
| 2 | Gemini 3.1 Pro (Preview) (High) | 69.78 |
| 3 | Kimi K2.5 (Thinking) | 65.5 |
| 4 | Gemini 3 Flash (Preview) (High) | 63.94 |
| 5 | DeepSeek V4 Pro (Max) | 63.24 |
| 6 | DeepSeek V4 Flash (Max) | 61.43 |
| 7 | DeepSeek V3.2 (Thinking) | 60.43 |
| 8 | MiMo-V2-Pro | 59.61 |
| 9 | MiniMax-M2.7 | 58.74 |
| 10 | GLM-5 | 58.32 |
| 11 | Tencent HY 2.0 Think | 57.58 |
| 12 | MiniMax-M2.5 | 55.33 |
| 13 | Grok 4.20 Beta (0309) (Reasoning) | 55.01 |
| 14 | MiMo-V2-Flash | 54.46 |
| 15 | GPT-OSS-120B (High) | 52.89 |
Interactive version: theaggregate.ai/benchmark?slug=superclue-general-march-2026-code-generation · How It Works · Data refreshed daily, snapshot 2026-09-19.