Answer Accuracy on BBox-DocVQA
90.92SPSBB AccuracyGPT-5.2
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| GPT-5.22026.06 | 90.92 | 93.53 | 91.51 | 91.92 | |
| Doubao-Seed-1-82026.06 | 89.98 | 90.11 | 89.94 | 90.02 | |
| Qwen3VLparameters=235B, variant=a22b-thinking2026.06 | 88.25 | 88.13 | 84.59 | 87.49 | |
| Qwen3.5parameters=27B2026.06 | 87.76 | 88.25 | 14.87 | 72.89 | |
| Gemini-3variant=Flash2026.06 | 85.31 | 78.96 | 68.55 | 79.85 | |
| GPT-52026.06 | 82.64 | 83.63 | 74.84 | 81.45 | |
| Qwen3VLparameters=32B2026.06 | 81.04 | 84.35 | 55.35 | 77.14 | |
| PreciseDoc-Reasoner2026.06 | 78.1 | 77.34 | 68.55 | 75.97 | |
| Gemini-3variant=Pro2026.06 | 74.36 | 46.4 | 39.62 | 57.98 | |
| Qwen2.5VLparameters=72B2026.06 | 71.03 | 71.58 | 57.86 | 68.64 | |
| Qwen3VLparameters=4B2026.06 | 70.49 | 72.12 | 59.75 | 68.95 | |
| Qwen3VLparameters=8B2026.06 | 70.23 | 73.2 | 51.57 | 67.59 | |
| Qwen2.5VLparameters=32B2026.06 | 67.69 | 66.55 | 48.11 | 63.46 | |
| Qwen2.5VLparameters=7B2026.06 | 57.01 | 53.42 | 35.53 | 51.57 | |
| InternVL3parameters=8B2026.06 | 53.14 | 53.06 | 39.62 | 50.46 | |
| InternVL3parameters=2B2026.06 | 39.12 | 33.27 | 24.21 | 34.2 | |
| Qwen2.5VLparameters=3B2026.06 | 31.38 | 33.27 | 25.47 | 30.87 |