Visual Text Reasoning and Recognition on OCRBench v2
76.7Recognition AccuracyUniDFlow
Evaluation results
| Method | Links | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|
| UniDFlowParameters=14B2026.02 | 76.7 | 47.1 | 16.5 | 96.9 | 48.4 | 58.1 | 88.7 | 77.1 | 63.8 | |
| UniDFlowParameters=8B2026.02 | 72.2 | 43.8 | 14.9 | 95 | 45.7 | 55.1 | 85.9 | 73.5 | 60.7 | |
| UniDFlowParameters=4B2026.02 | 69.9 | 41.2 | 12.9 | 94.1 | 42.2 | 53.4 | 83.1 | 70.8 | 58.4 | |
| Qwen2.5-VLParameters=7B2026.02 | 68.8 | 25.7 | 1.2 | 80.2 | 30.4 | 38.2 | 73.2 | 56.2 | 46.7 | |
| MammothModa22026.02 | 68.2 | 39.5 | 11.4 | 92.2 | 39.1 | 50.2 | 80.2 | 68.1 | 56.1 | |
| InternVL3Parameters=14B2026.02 | 67.3 | 36.9 | 11.2 | 89 | 38.4 | 38.4 | 79.2 | 60.5 | 52.6 | |
| Emma2026.02 | 66.7 | 36.5 | 6.7 | 91.3 | 37.5 | 44.5 | 76.7 | 67.2 | 53.8 | |
| Begal2026.02 | 65.8 | 37.1 | 3.3 | 90.45 | 38.5 | 41.3 | 75.2 | 66.4 | 52.2 | |
| Muddit2026.02 | 64.9 | 38.4 | 13.7 | 92.6 | 34.5 | 49.4 | 78.3 | 66.1 | 54.7 | |
| Qwen3-VLParameters=8B2026.02 | 64.4 | 38.2 | 5.7 | 91.03 | 37.8 | 44.2 | 76.8 | 62.6 | 55.7 | |
| OmniGenv22026.02 | 61.3 | 36.5 | 2.4 | 87.23 | 33.4 | 40.7 | 72.7 | 65.7 | 48.7 | |
| GPT-4o2026.02 | 61.2 | 26.7 | 0 | 77.5 | 36.3 | 43.4 | 71.1 | 55.5 | 46.5 | |
| Gemini-pro2026.02 | 61.2 | 39.5 | 13.5 | 79.3 | 39.2 | 47.7 | 75.5 | 59.3 | 51.9 | |
| GPT-4o-mini2026.02 | 57.9 | 23.3 | 0.6 | 70.8 | 31.5 | 38.8 | 65.9 | 55.1 | 43 |