Visual Perception on TreeBench
45.2ScoreMetis
Evaluation results
| Method | Links | |
|---|---|---|
| MetisBackbone=Qwen3-VL-8B-Instruct2026.04 | 45.2 | |
| Qwen2.5-VL-32B-InstructModel Category=Open-Source Models2026.04 | 42.5 | |
| DeepEyesV2Model Category=Agentic Multimodal Models2026.04 | 42.5 | |
| Qwen3-VL-8B-InstructModel Category=Open-Source Models2026.04 | 40.7 | |
| Pixel-ReasonerModel Category=Agentic Multimodal Models2026.04 | 39 | |
| InternVL3-8BModel Category=Open-Source Models2026.04 | 38.8 | |
| DeepEyesModel Category=Agentic Multimodal Models2026.04 | 37.5 | |
| LLaVA-OneVisionModel Category=Open-Source Models2026.04 | 37.3 | |
| Qwen2.5-VL-7B-InstructModel Category=Open-Source Models2026.04 | 37 |