M3-Bench - L1: leaderboard
Metric: Score. Source: huggingface.co. 36 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | DeepSeek V3.2 Speciale | 95.2 |
| 2 | GPT-5 Pro | 95.2 |
| 3 | GPT-5.1 | 93 |
| 4 | Gemini 3 Deep Think | 93 |
| 5 | Claude Opus 4.5 | 91 |
| 6 | Gemini 3 Pro | 89.7 |
| 7 | Grok 4.1 | 88.7 |
| 8 | GPT-OSS-120B | 87 |
| 9 | Kimi K2 (Thinking) | 87 |
| 10 | DeepSeek V3.2 | 86.7 |
| 11 | Mistral Large 3 | 81.8 |
Interactive version: theaggregate.ai/benchmark?slug=m3-bench-l1 · How It Works · Data refreshed daily, snapshot 2026-09-19.