Multimodal Perception and Reasoning on MME-RealWorld Lite
59.87Overall ScoreSCOLAR-7B
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| SCOLAR-7BBackbone=Qwen2.5-VL-7B2026.05 | 59.87 | 55.14 | 63.13 | |
| Monet2026.05 | 55.5 | 51.07 | 58.34 | |
| CapImaginebackbone=Qwen2.5-VL-7B2026.02 | 54.8 | 48.5 | 58.9 | |
| Deepeyes2026.05 | 54.28 | 50.53 | 56.63 | |
| SCOLARtraining_stage=SFT only2026.05 | 53.62 | 48 | 57.23 | |
| CapImaginebackbone=Qwen2.5-VL-7B, rewriting=false2026.02 | 53.5 | 43.7 | 57.6 | |
| DeepEyesbackbone=Qwen2.5-VL-7B2026.02 | 53.2 | 45.6 | 58.1 | |
| GPT-4o2026.05 | 52 | 48.3 | 54.4 | |
| GPT-4oreported_from_prior_work=true2026.02 | 52 | 48.3 | 54.4 | |
| LVRbackbone=Qwen2.5-VL-7B2026.02 | 50.6 | 42.7 | 55.7 | |
| GPT-4V2026.05 | 50.23 | 45.87 | 53.04 | |
| ActiveScopeModel=Qwen3-VL 4B2026.06 | 50.02 | 56.4 | 41.27 | |
| PixelReasonerbackbone=Qwen2.5-VL-7B2026.02 | 49.7 | 44.5 | 53.1 | |
| ZoomEyeModel=Qwen3-VL 4B2026.06 | 49.4 | 55.6 | 39.73 | |
| InternVL3-8Bbackbone=InternVL3-8B2026.02 | 47.9 | 42.9 | 51.1 | |
| Monetbackbone=Qwen2.5-VL-7B2026.02 | 46.9 | 40.3 | 51.2 | |
| Qwen3-VL 4BMode=Regular2026.06 | 46.74 | 52.01 | 38.53 | |
| GPT-4oMode=Regular2026.06 | 46.4 | 42.1 | 49.1 | |
| CapImaginebackbone=Qwen2.5-VL-7B, filtering=false2026.02 | 46.1 | 40.9 | 49.5 | |
| Qwen2.5VL-7Bbackbone=Qwen2.5VL-7B2026.02 | 45.8 | 39.7 | 49.6 | |
| Qwen2.5-VL-7B2026.05 | 45.75 | 39.73 | 49.62 |