Skip to content

Benchmark

Vdebench

Scores are percentages. Higher is better.

Models measured
9
Scores recorded
11
Sources
1
IQ weight
0.95

Score distribution

62%
Mean
63.8%
Median
12.8%
Std dev
45.9%
10th percentile
76.7%
90th percentile
42.3%
Min
42.3%80.6%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.95 for image-generation models, measured over 8 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
1Qwen Image 2.1· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim80.6%
2FLUX.2 [klein] 4B· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim76.7%
3Qwen Image Edit· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim75.7%
4Longcat Image Edit· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim70.1%
5Qwen Image Edit2511· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim65.6%
6firered-image-edit-v1.1· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim63.8%
7Qwen Image Edit2509· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim61.8%
8Hidream O1 Image· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim51%
9Step1X-Edit· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim48.2%
10GLM-Image· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim45.9%
11Flux Kontext Dev· Hugging Face Official Benchmarks data· Reported by the model's developer. The source relaying it did not verify it. claim42.3%

Data credits: Hugging Face official benchmark leaderboards; licence varies by benchmark