SuperCLUE General (March 2026) - Overall: leaderboard
Metric: Score. Source: www.superclueai.com. 24 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.6 (Max) | 77.02 |
| 2 | Gemini 3.1 Pro (Preview) (High) | 76.69 |
| 3 | GPT-5.4 (xHigh) | 72.48 |
| 4 | DeepSeek V4 Pro (Max) | 70.98 |
| 5 | Gemini 3 Flash (Preview) (High) | 68.84 |
| 6 | DeepSeek V4 Flash (Max) | 68.82 |
| 7 | Grok 4.20 Beta (0309) (Reasoning) | 66.07 |
| 8 | Kimi K2.5 (Thinking) | 64.6 |
| 9 | Qwen 3.5 397B A17B (Thinking) | 64.48 |
| 10 | GLM-5 | 64.27 |
| 11 | DeepSeek V3.2 (Thinking) | 61.92 |
| 12 | MiMo-V2-Pro | 60.67 |
| 13 | Tencent HY 2.0 Think | 59.16 |
| 14 | Qwen 3.5 122B A10B (Thinking) | 58.53 |
| 15 | LongCat-Flash-Thinking-2601 | 57.47 |
Interactive version: theaggregate.ai/benchmark?slug=superclue-general-march-2026-overall · How It Works · Data refreshed daily, snapshot 2026-09-19.