MMM-Bench - Level 3 (Business Stage): leaderboard
Metric: Macro-F1 (%) at taxonomy level 3 (Business Stage), zero-shot hierarchical classification of the test split (20 percent of 5,990 real Alibaba business documents in 12 domains, mostly Chinese) from OCR text, page images and filename against a flattened five-level taxonomy, temperature 0.1; higher is better. Source: arxiv.org. Saturation forecast: Around January 2027. 15 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Gemini 3.1 Pro (Preview) | 75.91 |
| 2 | Claude Opus 4.6 | 73.23 |
| 3 | Claude Opus 4.5 | 69.74 |
| 4 | Kimi K2.5 | 64.59 |
| 5 | Qwen 3.6 Plus | 64.51 |
| 6 | GLM-5 | 63.95 |
| 7 | GPT-5.4 | 62 |
| 8 | GPT-5.2 | 60.59 |
| 9 | MiniMax-M2.7 | 58.54 |
| 10 | Qwen 3.5 27B | 56.88 |
| 11 | Qwen 3 VL 235B A22B | 51.99 |
| 12 | Qwen 3.5 35B A3B | 49.53 |
| 13 | Qwen 3.5 9B | 41.02 |
| 14 | Qwen 2.5 VL 7B | 6.39 |
Interactive version: theaggregate.ai/benchmark?slug=mmm-bench-level-3-business-stage · How It Works · Data refreshed daily, snapshot 2026-10-07.