MuirBench
A comprehensive benchmark for robust multi-image understanding capabilities of multimodal LLMs. Consists of 12 diverse multi-image tasks involving 10 categories of multi-image relations (e.g., multiview, temporal relations, narrative, complementary). Comprises 11,264 images and 2,600 multiple-choice questions created in a pairwise manner, where each standard instance is paired with an unanswerable variant for reliable assessment.
11rows
scoreprimary metric
2026-05-06sampled
Metadata
Metrics
Score, Normalized Score
| Rank | Subject | Score | Model Match | Provenance | Sampled |
|---|---|---|---|---|---|
| 1 | Qwen3 VL 32B Thinking | 0.80 | — | Self-reported | 2026-05-06 |
| 2 | Qwen3 VL 235B A22B Thinking | 0.80 |  Qwen3 VL 235B A22B Thinking qwen-qwen3-vl-235b-a22b-thinking | Self-reported | 2026-05-06 |
| 3 | Qwen3 VL 30B A3B Thinking | 0.78 | Qwen3 VL 30B A3B Thinking qwen-qwen3-vl-30b-a3b-thinking | Self-reported | 2026-05-06 |
| 4 | Qwen3 VL 8B Thinking | 0.77 | Qwen3 VL 8B Thinking qwen-qwen3-vl-8b-thinking | Self-reported | 2026-05-06 |
| 5 | Qwen3 VL 4B Thinking | 0.75 | — | Self-reported | 2026-05-06 |
| 6 | Qwen3 VL 32B Instruct | 0.73 | Qwen3 VL 32B Instruct qwen-qwen3-vl-32b-instruct | Self-reported | 2026-05-06 |
| 6 | Qwen3 VL 235B A22B Instruct | 0.73 | Qwen3 VL 235B A22B Instruct qwen-qwen3-vl-235b-a22b-instruct | Self-reported | 2026-05-06 |
| 8 | Qwen3 VL 8B Instruct | 0.64 | Qwen3 VL 8B Instruct qwen-qwen3-vl-8b-instruct | Self-reported | 2026-05-06 |
| 9 | Qwen3 VL 4B Instruct | 0.64 | — | Self-reported | 2026-05-06 |
| 10 | Qwen3 VL 30B A3B Instruct | 0.63 | Qwen3 VL 30B A3B Instruct qwen-qwen3-vl-30b-a3b-instruct | Self-reported | 2026-05-06 |
| 11 | Qwen2.5-Omni-7B | 0.59 | — | Self-reported | 2026-05-06 |
No matching rows.