Our new X account is live! Follow @wizwand_team for updates

Multi-image reasoning on Muirbench (test)

68Accuracy

GPT-4V

Updated 3d ago

Evaluation Results

Method	Links
GPT-4V 2024.09		68
Qwen2.5VL 2025.12		59.6
MM1.5-30B 2024.09		58.2
MM1.5-7B 2024.09		49.1
LLaDA-V 2025.12		48.3
Qwen2.5VL 2025.12		47.7
DiffusionVL 2025.12		47.2
MM1.5-3B-MoE 2024.09		45.6
DiffusionVL 2025.12		44.8
Mantis-Idefics2-8B 2024.09		44.5
MM1.5-3B 2024.09		44.3
LLaVA-OV 2025.12		41.8
MM1.5-1B-MoE 2024.09		40.9
LLaVA-NeXT-Interleave-7B 2024.09		38.9
Phi-3-Vision-4B 2024.09		38
MM1-30B 2024.09		36.7
MM1.5-1B 2024.09		34.7
Emu2-Chat-37B 2024.09		33.6
MM1-1B 2024.09		30.7
MM1-7B 2024.09		30.4
MM1-3B 2024.09		28
Idefics2-8B 2024.09		26.1
LLaVAOneVision-0.5B 2024.09		25.5
LLaVA-v1.5-7B 2024.09		23.5