Mathematical Visual Question Answering on MathVision
73.3AccuracyGemini 2.5 Pro
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 2.5 Pro2026.04 | 73.3 | |
| OpenVLThinkerV2Parameters=8B2026.04 | 53.4 | |
| Qwen3-VL GRPORL Strategy=GRPO2026.04 | 52.6 | |
| ARES-7BParameters=7B2026.04 | 51.9 | |
| OVR-7BParameters=7B2026.04 | 51.8 | |
| Qwen3-VL GDPORL Strategy=GDPO2026.04 | 49.7 | |
| OneThinker-8BParameters=8B2026.04 | 48.3 | |
| Qwen3-VL-Instruct-8BParameters=8B2026.04 | 45.4 | |
| VL-Rethinker-7BParameters=7B2026.04 | 32.3 | |
| Vision-G12026.04 | 31.3 | |
| GPT-4o2026.04 | 30.4 | |
| VisionZero2026.04 | 28.5 | |
| MM-Eureka-7BParameters=7B2026.04 | 26.9 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=VQA2026.03 | 25.66 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=VQA2026.03 | 25.36 | |
| OpenVLThinker-7BParameters=7B2026.04 | 25.3 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=GQA2026.03 | 25.23 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.03 | 25.16 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.03 | 25.13 | |
| BaselineModel=Qwen2.5-7B-VL, Training=Inference Only2026.03 | 24.87 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-7B-VL, Train Set=GQA2026.03 | 24.54 | |
| SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=HiTab2026.03 | 22.76 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.03 | 18.78 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.03 | 18.55 | |
| SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=VQA2026.03 | 18.36 | |
| SFT (HQ Correct Only)Model=Qwen2.5-7B-VL, Train Set=GQA2026.03 | 18.09 | |
| BaselineModel=Qwen2.5-3B-VL, Training=Inference Only2026.03 | 17.76 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=GQA2026.03 | 17.5 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=VQA2026.03 | 17.4 | |
| VD-LB (HQ Correct vs. LQ Wrong)Model=Qwen2.5-3B-VL, Train Set=VQA2026.03 | 16.68 | |
| VD-LF (HQ vs. LQ)Model=Qwen2.5-3B-VL, Train Set=GQA2026.03 | 16.68 | |
| SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=HiTab2026.03 | 13.45 | |
| SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=VQA2026.03 | 11.97 | |
| SFT (HQ Correct Only)Model=Qwen2.5-3B-VL, Train Set=GQA2026.03 | 10.76 |