Reasoning and Math on VLMEvalKit (test)
73.4MathVista AccuracyGemini2.0-Flash
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| Gemini2.0-FlashModel Category=Proprietary Model2025.09 | 73.4 | 41.3 | 57.1 | 54.4 | 56.2 | 43.7 | 54.4 | |
| VisionZero-Qwen-7B (Real-World)Backbone=Qwen2.5-VL-7B, Training Domain=Real-World2025.09 | 73.1 | 28.5 | 40.1 | 52.1 | 50.8 | 22.5 | 44.5 | |
| MM-Eureka-Qwen-7BBackbone=Qwen2.5-VL-7B2025.09 | 73 | 26.9 | 36.2 | 50.3 | 42.9 | 24.2 | 42.9 | |
| VisionZero-Qwen-7B (Chart)Backbone=Qwen2.5-VL-7B, Training Domain=Chart2025.09 | 72.4 | 28.9 | 38.8 | 52.5 | 51.2 | 22.1 | 43.3 | |
| VisionZero-Qwen-7B (CLEVR)Backbone=Qwen2.5-VL-7B, Training Domain=CLEVR2025.09 | 72.2 | 28.4 | 39.2 | 53.2 | 49.8 | 22.9 | 44.3 | |
| ViGaL-Snake+RotationBackbone=Qwen2.5-VL-7B2025.09 | 71.9 | 27.5 | 36.9 | 52.4 | 46.5 | 22.9 | 43 | |
| ViGaL-RotationBackbone=Qwen2.5-VL-7B2025.09 | 71.2 | 26.3 | — | 50.4 | — | — | — | |
| ViGaL-SnakeBackbone=Qwen2.5-VL-7B2025.09 | 70.7 | 26.5 | — | 51.1 | — | — | — | |
| OpenVLThinker-7BBackbone=Qwen2.5-VL-7B2025.09 | 70.2 | 25.3 | 36.5 | 47.9 | 44.3 | 21.2 | 40.9 | |
| Qwen2.5-VL-7BBackbone=Qwen2.5-VL-7B2025.09 | 68.2 | 25.4 | 36.1 | 49 | 47.2 | 20.9 | 41.1 | |
| VLAA-Thinker-7BBackbone=Qwen2.5-VL-7B2025.09 | 68 | 26.4 | 36 | 51.7 | 47.2 | 21.9 | 41.9 | |
| R1-OneVision-7BBackbone=Qwen2.5-VL-7B2025.09 | 64.1 | 24.1 | 35.8 | 47.1 | 44.5 | 21.4 | 39.5 | |
| GPT4oModel Category=Proprietary Model2025.09 | 61.4 | 30.4 | 40 | 50.2 | 45.9 | 32.3 | 43.4 |