Visual Perception on MME-RealWorld
71.4ScoreSkywork-R1V4-30B-A3B
Evaluation Results
| Method | Links | |
|---|---|---|
| Skywork-R1V4-30B-A3BModel Category=Agentic Multimodal Models2026.04 | 71.4 | |
| MetisBackbone=Qwen3-VL-8B-Instruct2026.04 | 70.3 | |
| SenseNova-MARS-8BModel Category=Agentic Multimodal Models2026.04 | 67.9 | |
| Mini o3Model Category=Agentic Multimodal Models2026.04 | 65.5 | |
| DeepEyesV2Model Category=Agentic Multimodal Models2026.04 | 64.9 | |
| ThymeModel Category=Agentic Multimodal Models2026.04 | 64.8 | |
| Pixel-ReasonerModel Category=Agentic Multimodal Models2026.04 | 64.4 | |
| DeepEyesModel Category=Agentic Multimodal Models2026.04 | 64.1 | |
| Qwen3-VL-8B-InstructModel Category=Open-Source Models2026.04 | 61.9 | |
| Qwen2.5-VL-32B-InstructModel Category=Open-Source Models2026.04 | 59.1 | |
| LLaVA-OneVisionModel Category=Open-Source Models2026.04 | 57.4 | |
| Qwen2.5-VL-7B-InstructModel Category=Open-Source Models2026.04 | 56.8 |