Skip to content

Benchmark

DocVQAtest

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

Models measured
7
Scores recorded
9
Sources
1
IQ weight
0.32

Score distribution

96.1%
Mean
96.1%
Median
0.7%
Std dev
95%
10th percentile
96.9%
90th percentile
95%
Min
95%97.1%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.32 for chat models, measured over 7 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 VL 235B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.97.1%
2Qwen3 VL 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.96.9%
3Qwen2-VL-72B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim96.5%
4Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.96.5%
5Qwen3 VL 32B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.96.1%
6Qwen3 VL 8B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.96.1%
7Qwen: Qwen3 VL 8B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.95.3%
8qwen3-vl-30b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.95%
9qwen3-vl-30bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.95%

Data credits: Data by LLM Stats