Skip to content

Benchmark

MMBenchEN-DEV-v1.1

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

Models measured
5
Scores recorded
5
Sources
1
IQ weight
1.33

Score distribution

92.1%
Mean
92.8%
Median
1.7%
Std dev
88.8%
10th percentile
93.7%
90th percentile
88.8%
Min
88.8%93.7%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 1.33 for chat models, measured over 5 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen3.5 397B A17BQwen3.5 release/model card · published setting· Reported by Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.93.7%
2Qwen3.5 Omni PlusQwen3.5-Omni release · published setting· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.92.8%
3Qwen3.6 FlashQwen3.6-35B-A3B official model card · hosted API name Qwen3.6-Flash · DEV v1.1· Reported by Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.92.8%
4Qwen3.6 27BQwen official release table · thinking mode unless benchmark requires otherwise · DEV v1.1· Reported by Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.92.3%
5Qwen3.5 Omni FlashQwen3.5-Omni release · published setting· Reported by Qwen, Qwen Team· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.88.8%

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.