MMBench-Live - Fine-grained Perception (Cross Instance): leaderboard
Metric: Accuracy (%; cross-instance fine-grained perception questions; 5.9K newly generated evaluation instances aligned with the MMBench task hierarchy; temperature 0). Source: arxiv.org. Saturation forecast: Around December 2026. 6 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Qwen 2.5 VL 7B Instruct | 77.47 |
| 2 | Qwen 3 VL 8B Instruct | 70.72 |
Interactive version: theaggregate.ai/benchmark?slug=mmbench-live-fine-grained-perception-cross-instance · How It Works · Data refreshed daily, snapshot 2026-09-29.