General Task on MMStar (accuracy)
76.2AccuracySeed1.5-VL
Evaluation Results
| Method | Links | |
|---|---|---|
| Seed1.5-VLModel=Seed1.5-VL, Training Strategy=Reference2025.10 | 76.2 | |
| Qwen2.5-VL-72B-InstructModel=Qwen2.5-VL-72B-Instruct, Training Strategy=Reference2025.10 | 70.8 | |
| Qwen2.5-VL-32B-InstructModel=Qwen2.5-VL-32B-Instruct, Training Strategy=Reference2025.10 | 69.5 | |
| InternVL3-8BModel Category=Open-source Models2026.05 | 68.2 | |
| Thyme-7BModel Category=Tool-based Models2026.05 | 65.9 | |
| DeepLatent-RL-7BModel Category=Latent Visual reasoning Models, Training Stage=RL, Training Data=full DeepLatent-180K dataset2026.05 | 65 | |
| DeepLatent-RL-7B*Model Category=Latent Visual reasoning Models, Training Stage=RL, Training Data=visual search data of DeepLatent-180K in SFT Stage 22026.05 | 64.1 | |
| DeepLatent-SFT-7BModel Category=Latent Visual reasoning Models, Training Stage=SFT2026.05 | 63.6 | |
| AODModel=Qwen2.5-VL-7B2026.05 | 63 | |
| Qwen2.5-VL-7B-Instruct + SFT -> RLModel=Qwen2.5-VL-7B-Instruct, Training Strategy=SFT -> RL Baseline2025.10 | 62.27 | |
| REWARDMAPModel=Qwen2.5-VL-7B-Instruct, Training Strategy=REWARDMAP2025.10 | 62.27 | |
| LLaVA-OneVision-7BModel Category=Open-source Models2026.05 | 61.9 | |
| Qwen2.5-VL-7BModel Category=Open-source Models2026.05 | 61.9 | |
| Kimi-VL-A3B-InstructModel=Kimi-VL-A3B-Instruct, Training Strategy=Reference2025.10 | 61.7 | |
| Qwen2.5-VL-7B-InstructModel=Qwen2.5-VL-7B-Instruct, Training Strategy=Base Model2025.10 | 61.67 | |
| TruthPrIntModel=Qwen2.5-VL-7B2026.05 | 61.1 | |
| ASDModel=Qwen2.5-VL-7B2026.05 | 60.7 | |
| AODModel=InternVL3-8B2026.05 | 60.2 | |
| VASparseModel=Qwen2.5-VL-7B2026.05 | 60.1 | |
| VCDModel=Qwen2.5-VL-7B2026.05 | 59.6 | |
| PruneHalModel=InternVL3-8B2026.05 | 59 | |
| BaseModel=Qwen2.5-VL-7B2026.05 | 58.9 | |
| PruneHalModel=Qwen2.5-VL-7B2026.05 | 57.8 | |
| TruthPrIntModel=InternVL3-8B2026.05 | 56.8 | |
| ASDModel=InternVL3-8B2026.05 | 56.1 | |
| VASparseModel=InternVL3-8B2026.05 | 56 | |
| Qwen2.5-VL-3B-InstructModel=Qwen2.5-VL-3B-Instruct, Training Strategy=Reference2025.10 | 55.9 | |
| BaseModel=InternVL3-8B2026.05 | 55.8 | |
| VCDModel=InternVL3-8B2026.05 | 54.6 | |
| AODModel=LLaVA-1.5-7B2026.05 | 36.3 | |
| ASDModel=LLaVA-1.5-7B2026.05 | 35.8 | |
| TruthPrIntModel=LLaVA-1.5-7B2026.05 | 35.2 | |
| PruneHalModel=LLaVA-1.5-7B2026.05 | 35.1 | |
| VCDModel=LLaVA-1.5-7B2026.05 | 34.1 | |
| VASparseModel=LLaVA-1.5-7B2026.05 | 33.5 | |
| BaseModel=LLaVA-1.5-7B2026.05 | 32.6 |