Skip to content

Benchmark

MMBench

Vision benchmarks measure how well models understand images. Scores are percentages.

Models measured
6
Scores recorded
6
Sources
1
IQ weight
0.39

Score distribution

84.5%
Mean
86.5%
Median
2.9%
Std dev
81.9%
10th percentile
88%
90th percentile
79.6%
Min
79.6%88%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.39 for chat models, measured over 6 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen2.5 VL 72B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim88%
2Phi 4 Multimodal Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim86.7%
3Qwen2-VL-72B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim86.5%
4Qwen2.5-VL 7B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim84.3%
5Phi-3.5 Vision Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim81.9%
6DeepSeek VL2· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim79.6%

Data credits: Data by LLM Stats