Visual perception of small objects on V* Bench
94.76AccuracyVision-OPD (Ours)
Evaluation Results
| Method | Links | |
|---|---|---|
| Vision-OPD (Ours)Param Size=9B, Inference Mode=Single Forward Pass2026.05 | 94.76 | |
| SenseNova-MARSParam Size=8B, Inference Mode=Agentic2026.05 | 92.15 | |
| Vision-OPD (Ours)Param Size=4B, Inference Mode=Single Forward Pass2026.05 | 92.15 | |
| Qwen3-VL-InstructParam Size=235B, Inference Mode=Single Forward Pass2026.05 | 91.1 | |
| Gemini-3.1-ProParam Size=-, Inference Mode=Closed-Source2026.05 | 87.96 | |
| ZwZParam Size=8B, Inference Mode=Single Forward Pass2026.05 | 87.96 | |
| Qwen3.5Param Size=397B, Inference Mode=Single Forward Pass2026.05 | 87.96 | |
| Kimi-K2.5Param Size=1T, Inference Mode=Single Forward Pass2026.05 | 87.43 | |
| GLM-4.6VParam Size=106B, Inference Mode=Single Forward Pass2026.05 | 86.91 | |
| DeepEyesParam Size=7B, Inference Mode=Agentic2026.05 | 85.86 | |
| Qwen3-VL-InstructParam Size=8B, Inference Mode=Single Forward Pass2026.05 | 84.82 | |
| Qwen3.5Param Size=4B, Inference Mode=Single Forward Pass2026.05 | 84.29 | |
| MiMo-VL-RLParam Size=7B, Inference Mode=Single Forward Pass2026.05 | 83.25 | |
| Qwen3.5Param Size=9B, Inference Mode=Single Forward Pass2026.05 | 82.72 | |
| ThymeParam Size=7B, Inference Mode=Agentic2026.05 | 82.2 | |
| DeepEyesV2Param Size=7B, Inference Mode=Agentic2026.05 | 81.68 | |
| GPT-5.2Param Size=-, Inference Mode=Closed-Source2026.05 | 79.06 | |
| GPT-5.4Param Size=-, Inference Mode=Closed-Source2026.05 | 76.96 | |
| MiniCPM-V-4.5Param Size=9B, Inference Mode=Single Forward Pass2026.05 | 70.68 |