MMBU - Grounded Classification (Masks): leaderboard
Metric: Closed-ended classification of a region marked by a segmentation mask: micro-averaged F1 x100 (0-100) over MMBU's biomedical images from 35 submodalities (95% bootstrap intervals not published); zero-shot, official checkpoints; exact match after normalization. Source: arxiv.org. Saturation forecast: Around December 2026. 17 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 VL 32B Instruct | 69.3 |
| 2 | Qwen 3 VL 32B Instruct | 63 |
| 3 | Qwen 3 VL 4B Instruct | 54 |
| 4 | Qwen 3 VL 8B Instruct | 52.6 |
| 5 | InternVL3.5-8B | 51.4 |
| 6 | GPT-4.1 Mini | 50.1 |
| 7 | GPT-5.4 Mini | 47.2 |
| 8 | Gemma 3 4B (IT) | 38.9 |
| 9 | Qwen 2.5 VL 7B Instruct | 35.7 |
| 10 | MedGemma-4B-IT | 31 |
| 11 | Lingshu-32B | 31 |
| 12 | Lingshu-7B | 25.8 |
Interactive version: theaggregate.ai/benchmark?slug=mmbu-grounded-classification-masks · How It Works · Data refreshed daily, snapshot 2026-09-29.