Benchmark
OmniDocBench - Formula CDM
OCR benchmarks measure how well models read text in images and documents. Scores are percentages.
- Models measured
- 11
- Scores recorded
- 11
- Sources
- 1
- IQ weight
- 0.22
Score distribution
- 92.8%
- Mean
- 95.1%
- Median
- 4.3%
- Std dev
- 87.7%
- 10th percentile
- 97.2%
- 90th percentile
- 83.5%
- Min
This benchmark has a weight of 0.22 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
| # | Model | Score |
|---|---|---|
| 1 | PaddleOCR-VL-1.6· OmniDocBench data | 97.5% |
| 2 | GLM OCR· OmniDocBench data | 97.2% |
| 3 | Gemini 3 Pro Preview· OmniDocBench data | 96% |
| 4 | PaddleOCR VL· OmniDocBench data | 95.9% |
| 5 | Gemini 3 Flash· OmniDocBench data | 95.2% |
| 6 | Qianfan-OCR· OmniDocBench data | 95.1% |
| 7 | Qwen3 VL 235B· OmniDocBench data | 92.6% |
| 8 | DeepSeek OCR 2· OmniDocBench data | 91.8% |
| 9 | GPT 5.2· OmniDocBench data | 88.2% |
| 10 | HunyuanOCR· OmniDocBench data | 87.7% |
| 11 | Kimi K2.5· OmniDocBench data | 83.5% |
Data credits: OmniDocBench (OpenDataLab), Apache-2.0 (licence)