List your company
ResearchFormerly WizWand

Real-world Multimodal Understanding on MME-RealWorld Lite

54.9Lite ScoreZwZ-Qwen3-VL-4B + Q-Zoom
26.50833.87941.2548.621Apr 8, 2026
Updated 6mo ago

Evaluation results

MethodLinks
ZwZ-Qwen3-VL-4B + Q-ZoomRelative inference throughput (Tp)=0.66×2026.0454.9
ZwZ-Qwen3-VL-4BRelative inference throughput (Tp)=1.00×2026.0454.3
Qwen2.5-VL-7B + ThymeRelative inference throughput (Tp)=0.21×2026.0453.7
Qwen3-VL-4B + Q-ZoomRelative inference throughput (Tp)=0.73×2026.0453.5
ZwZ-Qwen2.5-VL-7B + Q-ZoomRelative inference throughput (Tp)=0.76×2026.0453.2
ZwZ-Qwen2.5-VL-7BRelative inference throughput (Tp)=1.00×2026.0452.5
Qwen3-VL-4B + SD-RPNRelative inference throughput (Tp)=0.58×2026.0451.8
Qwen2.5-VL-7B + DeepEyesEvaluation protocol=Directly cited (†)2026.0450.9
Qwen2.5-VL-7B + Q-ZoomRelative inference throughput (Tp)=0.86×2026.0448
Qwen3-VL-4BRelative inference throughput (Tp)=1.00×2026.0447.4
Qwen2.5-VL-7B + SD-RPNRelative inference throughput (Tp)=0.77×2026.0446.4
Qwen2.5-VL-3B + Q-ZoomRelative inference throughput (Tp)=0.67×2026.0444
Qwen2.5-VL-7BRelative inference throughput (Tp)=1.00×2026.0442.7
Qwen2.5-VL-3B + SD-RPNRelative inference throughput (Tp)=0.66×2026.0441.9
Qwen2.5-VL-3BRelative inference throughput (Tp)=1.00×2026.0441.6
LLaVA-1.5-13B + S2Relative inference throughput (Tp)=0.73×2026.0436.3
LLaVA-1.5-13B + SD-RPNRelative inference throughput (Tp)=0.57×2026.0431.4
LLaVA-1.5-13B + Q-ZoomRelative inference throughput (Tp)=0.58×2026.0431.3
LLaVA-1.5-13B + ViCropRelative inference throughput (Tp)=0.21×2026.0430.1
LLaVA-1.5-7B + S2Relative inference throughput (Tp)=0.63×2026.0428.5
LLaVA-1.5-13BRelative inference throughput (Tp)=1.00×2026.0427.8
LLaVA-1.5-7BRelative inference throughput (Tp)=1.00×2026.0427.7
LLaVA-1.5-7B + SD-RPNRelative inference throughput (Tp)=0.57×2026.0427.7
LLaVA-1.5-7B + Q-ZoomRelative inference throughput (Tp)=0.58×2026.0427.7
LLaVA-1.5-7B + ViCropRelative inference throughput (Tp)=0.15×2026.0427.6