Slide Question Answering on SlideVQA
79.5Overall ScoreQwen2.5-VL-7B-Instruct
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| Qwen2.5-VL-7B-InstructType=12025.10 | 79.5 | 70.5 | 94.3 | — | — | |
| Qwen2.5-VL-32B-InstructType=12025.10 | 79.2 | 71.1 | 92.2 | — | — | |
| SlideAgentType=3, Base Model=InternVL3-8B2025.10 | 72.7 | 68.2 | 79.4 | — | — | |
| Phi-3-vision-128k-instructType=12025.10 | 72.3 | 61.8 | 90.6 | — | — | |
| ViDoRAGType=2, Base Model=InternVL3-8B2025.10 | 68.8 | 61.9 | 77.3 | — | — | |
| VDocRAGType=2, Base Model=InternVL3-8B2025.10 | 65.2 | 59.7 | 77 | — | — | |
| VisRAGType=2, Base Model=InternVL3-8B2025.10 | 63.6 | 56.5 | 75.5 | — | — | |
| COLPALIType=2, Base Model=InternVL3-8B2025.10 | 63.4 | 56.7 | 73.8 | — | — | |
| InternVL3-8BType=12025.10 | 63 | 56.5 | 74.1 | — | — | |
| VL-Scaler-MiMOSize=7B2026.05 | 59.6 | — | — | — | — | |
| VL-ScalerSize=7B2026.05 | 58.3 | — | — | — | — | |
| GPT-4oSize=-2026.05 | 53.1 | — | — | — | — | |
| llava-v1.6-mistral-7b-hfType=12025.10 | 50.9 | 37.3 | 82.6 | — | — | |
| mPLUG-Owl3Size=7B2026.05 | 48.8 | — | — | — | — | |
| GPT-4o-miniSize=-2026.05 | 48.1 | — | — | — | — | |
| MiMO-VL-InstructSize=7B2026.05 | 48 | — | — | — | — | |
| llava-1.5-13b-hfType=12025.10 | 44.9 | 25.1 | 81.8 | — | — | |
| Llama-3.2-11B-Vision-InstructType=12025.10 | 42.9 | 43.3 | 42.3 | — | — | |
| Pixel ReasonerSize=7B2026.05 | 42.1 | — | — | — | — | |
| Qwen2.5-VL-InstructSize=7B2026.05 | 38.9 | — | — | — | — | |
| Qwen2.5-VL-InstructSize=72B2026.05 | 38.8 | — | — | — | — | |
| VL-RethinkerSize=7B2026.05 | 38.5 | — | — | — | — | |
| llava-1.5-7b-hfType=12025.10 | 36.8 | 22.4 | 79 | — | — | |
| DocopilotSize=8B2026.05 | 35.7 | — | — | — | — | |
| Llava-OVSize=7B2026.05 | 35 | — | — | — | — | |
| OpenVLThinkerSize=7B2026.05 | 33.5 | — | — | — | — | |
| DeepEyesSize=7B2026.05 | 33 | — | — | — | — | |
| R1-VLSize=7B2026.05 | 22.1 | — | — | — | — | |
| llava-v1.6-vicuna-13b-hfType=12025.10 | 16.7 | 10.2 | 81.5 | — | — | |
| AVIR2026.04 | — | — | — | 60.3 | — | |
| Eagle-2.52026.04 | — | — | — | 63.2 | — | |
| Human2026.04 | — | — | — | 89.8 | — | |
| M3D2026.04 | — | — | — | 33.5 | — | |
| Qwen2-VL-7B-InstructBackbone=Qwen2-VL-7B-Instruct2026.04 | — | — | — | 28.5 | — | |
| Qwen2-VL-7B-Instruct + BEFTBackbone=Qwen2-VL-7B-Instruct, Training Protocol=BEFT2026.04 | — | — | — | 69.6 | — | |
| Qwen2-VL-7B-Instruct + SFTBackbone=Qwen2-VL-7B-Instruct, Training Protocol=SFT2026.04 | — | — | — | 31 | — | |
| VDocRAG2026.04 | — | — | — | — | 44.2 |