List your company
ResearchFormerly WizWand

High-resolution Perception on HR-Bench 8K

82ScoreMetis
53.460.82568.2575.675Mar 6, 2026Mar 26, 2026Apr 16, 2026May 6, 2026May 27, 2026Jun 16, 2026Jul 7, 2026
Updated 3mo ago

Evaluation results

MethodLinks
MetisBackbone=Qwen3-VL-8B-Instruct2026.0482
Skywork-R1V4-30B-A3BModel Category=Agentic Multimodal Models2026.0479.8
SenseNova-MARS-8BModel Category=Agentic Multimodal Models2026.0478.4
Qwen3-VL-8B-InstructModel Category=Open-Source Models2026.0474.6
DeepLatent-RL-7B*Model Category=Latent Visual reasoning Models, Training Stage=RL, Training Data=visual search data of DeepLatent-180K in SFT Stage 22026.0574.1
DeepLatent-RL-7BModel Category=Latent Visual reasoning Models, Training Stage=RL, Training Data=full DeepLatent-180K dataset2026.0574
MiMo-VL-7B +PCBackbone Model=MiMo-VL-7B, Training Paradigm (+PC)=Yes2026.0373.8
DeepEyesV2Model Category=Agentic Multimodal Models2026.0473.8
DeepLatent-SFT-7BModel Category=Latent Visual reasoning Models, Training Stage=SFT2026.0573.6
Mini o3Model Category=Agentic Multimodal Models2026.0473.3
DeepEyes-7BModel Category=Tool-based Models2026.0572.6
ThymeModel Category=Agentic Multimodal Models2026.0472
Thyme-7BModel Category=Tool-based Models2026.0572
SegAnswerBase Model=Qwen2.5-VL-7B2026.0771.3
MiMo-VL-7BBackbone Model=MiMo-VL-7B, Training Paradigm (+PC)=No2026.0370.6
CoVT-7BModel Category=Latent Visual reasoning Models2026.0569.9
DyFo-7BModel Category=Tool-based Models2026.0569.8
DeepEyesEvaluation protocol=Re-evaluated using its official model and evaluation code2026.0769.8
Qwen2.5-VL-7B +PCBackbone Model=Qwen2.5-VL-7B, Training Paradigm (+PC)=Yes2026.0369.6
DeepEyesModel Category=Agentic Multimodal Models2026.0469.5
InternVL3-8BModel Category=Open-Source Models2026.0469.3
InternVL3-8BModel Category=Open-source Models2026.0569.3
Monet-7BModel Category=Latent Visual reasoning Models2026.0568
Pixel-Reasoner-7BModel Category=Tool-based Models2026.0566.9
Pixel ReasonerEvaluation protocol=Re-evaluated using its official model and evaluation code2026.0766.4
Pixel-ReasonerModel Category=Agentic Multimodal Models2026.0466.1
Qwen2.5-VL-7BBackbone Model=Qwen2.5-VL-7B, Training Paradigm (+PC)=No2026.0365.3
Qwen2.5-VL-7BModel Category=Open-source Models2026.0565.3
Qwen2.5-VL-3B +PCBackbone Model=Qwen2.5-VL-3B, Training Paradigm (+PC)=Yes2026.0363.7
Qwen2.5-VL-32B-InstructModel Category=Open-Source Models2026.0463.6
Qwen2.5-VL-3BBackbone Model=Qwen2.5-VL-3B, Training Paradigm (+PC)=No2026.0363.5
Qwen2.5-VL-7BEvaluation protocol=Re-evaluated using its official model and evaluation code2026.0763.4
LVR-7BModel Category=Latent Visual reasoning Models2026.0563
Qwen2.5-VL-7B-InstructModel Category=Open-Source Models2026.0462.1
LLaVA-OneVisionModel Category=Open-Source Models2026.0459.8
LLaVA-OneVision-7BModel Category=Open-source Models2026.0559.8
LLaVA-OneVision-9BEvaluation protocol=Re-evaluated using its official model and evaluation code2026.0754.5