Skip to content

Benchmark

OCRBench-V2 (en)

OCR benchmarks measure how well models read text in images and documents. Scores are percentages.

Models measured
8
Scores recorded
10
Sources
1
IQ weight
0.79

Score distribution

65.4%
Mean
66.8%
Median
2.4%
Std dev
62.6%
10th percentile
68.1%
90th percentile
61.5%
Min
61.5%68.4%
The band covers the middle 80% of scores. The line marks the median.

This benchmark has a weight of 0.79 for chat 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
1Qwen3 VL 32B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.68.4%
2NVIDIA Nemotron 3.5 Super VL 120B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim68.1%
3Qwen3 VL 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.67.4%
4Qwen3 VL 235B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.67.1%
5Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.66.8%
6Qwen3 VL 8B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.65.4%
7Qwen: Qwen3 VL 8B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.63.9%
8qwen3-vl-30b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.63.2%
9qwen3-vl-30bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.62.6%
10Qwen2.5 VL 72B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim61.5%

Data credits: Data by LLM Stats