Benchmark

MuirBench

A comprehensive benchmark for robust multi-image understanding capabilities of multimodal LLMs. Consists of 12 diverse multi-image tasks involving 10…

Modality
multimodal
Categories
multimodal, reasoning, vision
Openness
unknown
Source
llm_stats
Reported scores
12

Reported scores

llm_stats

ModelOrganizationReported valueReportedEvidence
LFM2.5-VL-3BLiquid AI0.5832026-08-12evidence
Qwen2.5-Omni-7BQwen0.5922025-03-27evidence
Qwen3 VL 235B A22B InstructQwen0.7282025-09-22evidence
Qwen3 VL 235B A22B ThinkingQwen0.8012025-09-22evidence
Qwen3 VL 30B A3B InstructQwen0.6292025-09-22evidence
Qwen3 VL 30B A3B ThinkingQwen0.7762025-09-22evidence
Qwen3 VL 32B InstructQwen0.7282025-09-22evidence
Qwen3 VL 32B ThinkingQwen0.8032025-09-22evidence
Qwen3 VL 4B InstructQwen0.6382025-09-22evidence
Qwen3 VL 4B ThinkingQwen0.752025-09-22evidence
Qwen3 VL 8B InstructQwen0.6442025-09-22evidence
Qwen3 VL 8B ThinkingQwen0.7682025-09-22evidence

Scores are partitioned by the source that reported them and are never merged into a single cross-source ranking, because the sources measure different things and say so.