Skip to content

Benchmark

OCRBench

OCR benchmarks measure how well models read text in images and documents. Scores are percentages, except where a source reports on a different scale. Check each score's scale before comparing. Higher is better.

Models measured
64
Scores recorded
68
Sources
3
IQ weight
0.00

Score distribution

528.4%
Mean
701%
Median
358%
Std dev
0.877%
10th percentile
862%
90th percentile
0.811%
Min

This benchmark has a weight of 0.00 for chat models, measured over 64 models. The weight is how much it counts toward llminfo IQ. We measure it from the data, not from the raw scores.

Leaderboard

#ModelScore
51Qwen3.5 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.93.1%
52Kimi K2.5· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.3%
53Kimi K2.5setting not named by the source· Reported by Moonshot AI· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.92.3%
54Qwen3.5 122B-A10B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim92.1%
55Qwen3 VL 235B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.92%
56Qwen3.5 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.91.3%
57Qwen3.5 35B A3B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim91%
58qwen3-vl-30b· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.90.3%
59Qwen3 VL 8B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.89.6%
60Qwen3 VL 32B Instruct· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.89.5%
61Qwen3.5 27B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim89.4%
62Qwen3.6 27B· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it. claim89.4%
63Qwen3.6 27BQwen official release table · thinking mode unless benchmark requires otherwise· Reported by Alibaba Qwen· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.89.4%
64Qwen3.5 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.89.1%
65Qwen3 VL 235Bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.87.5%
66Qwen3 VL 32B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.85.5%
67qwen3-vl-30bthinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.83.9%
68Qwen: Qwen3 VL 8B Thinking· LLM Stats data· Reported by the model's developer. The source relaying it did not verify it.81.9%

Showing 51–68 of 68 results.

Data credits: OpenVLM Leaderboard (OpenCompass / VLMEvalKit), Apache-2.0 (licence) · Data by LLM Stats · 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.