OCRBench
AI benchmark results and model performance
Type: percentageImage To Text
Recorded results
50
Average score
559%
Score range
80.90% - 920%
Leading model
920% - Qwen 3 VL 235B A22B Instruct
Scores over time
Individual benchmark scores plotted by date.
Model results
| Organisation | Models | Reported | Top score | Info | Self reported | Source |
|---|---|---|---|---|---|---|
| Qwen 3 VL 235B A22B Instruct | - | 920% | - | Yes | - | |
| Qwen 3 VL 30B A3B Instruct | - | 903% | - | Yes | - | |
| Qwen 3 VL 8B Instruct | - | 896% | - | Yes | - | |
| Qwen 3 VL 32B Instruct | - | 895% | - | Yes | - | |
| Qwen 3 VL 4B Instruct | - | 881% | - | Yes | - | |
| Qwen 3 VL 235B A22B Thinking | - | 875% | - | Yes | - | |
| Qwen 3 VL 32B Thinking | - | 855% | - | Yes | - | |
| Qwen 3 VL 30B A3B Thinking | - | 839% | - | Yes | - | |
| Qwen 3 8B | - | 819% | inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.4600; benches=50) | Yes | - | |
| Qwen 3 Guard Stream 8B | - | 819% | inferred family alias from qwen3-vl-8b-thinking (score=0.3371; benches=50) | Yes | - | |
| Qwen 3 VL 8B Thinking | - | 819% | - | Yes | - | |
| Qwen 3 VL Embedding 8B | - | 819% | inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5232; benches=50) | Yes | - | |
| Qwen 3 VL Reranker 8B | - | 819% | inferred high-confidence family alias from qwen3-vl-8b-thinking (score=0.5275; benches=50) | Yes | - | |
| Qwen 3 4B Instruct 2507 | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48) | Yes | - | |
| Qwen 3 4B SafeRL | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48) | Yes | - | |
| Qwen 3 4B Thinking 2507 | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3462; benches=48) | Yes | - | |
| Qwen 3 Embedding 4B | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48) | Yes | - | |
| Qwen 3 Guard Gen 4B | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3400; benches=48) | Yes | - | |
| Qwen 3 Guard Stream 4B | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3371; benches=48) | Yes | - | |
| Qwen 3 Reranker 4B | - | 808% | inferred family alias from qwen3-vl-4b-thinking (score=0.3850; benches=48) | Yes | - | |
| Qwen 3 VL 4B Thinking | - | 808% | - | Yes | - | |
| Kimi K2.5 | 27 Jan 2026 | 92.30% | - | Yes | Source | |
| Qwen 3.5 122B A10B | 24 Feb 2026 | 92.10% | Raw score: 92.1 | Yes | Source | |
| Qwen 3.5 35B A3B | 24 Feb 2026 | 91% | Raw score: 91.0 | Yes | Source | |
| Qwen 3.5 27B | 24 Feb 2026 | 89.40% | Raw score: 89.4 | Yes | Source |