List your company
ResearchFormerly WizWand

Multimodal Understanding and Reasoning on MMBench Chinese (dev)

72.8AccuracyDeepseek-VL-7B
35.77645.3885564.612Jan 28, 2026
Updated 8mo ago

Evaluation results

MethodLinks
Deepseek-VL-7BLLM=DLLM-7B, #Sample=2000M, #Param=≥7B2026.0172.8
LLaVA-FA-7BLLM=InternLM-2-20B, #Sample=5M, #Param=≥7B2026.0169.5
LLaVA-FA-3BLLM=LLaMA-3-8B, #Sample=5M, #Param=∼3B2026.0168
MiniCPM-V-2LLM=MiniCPM-2.4B, #Sample=570M, #Param=∼3B2026.0167.2
LLaVA-NeXTLLM=Vicuna-1.5-13B, #Sample=1.3M, #Param=≥7B2026.0164.4
MiniCPM-VLLM=MiniCPM-2.4B, #Sample=570M, #Param=∼3B2026.0162.7
VILA-7BLLM=LLaMA-7B, #Sample=50M, #Param=≥7B2026.0161.7
LLaVA-FA-2BLLM=Qwen-2.5-7B, #Sample=5M, #Param=∼2B2026.0161.7
Imp-2BLLM=Qwen-1.5-1.8B, #Sample=1.6M, #Param=∼2B2026.0161.2
DeepSeek-VL-1.3BLLM=DLLM-1.3B, #Sample=2000M, #Param=∼2B2026.0161
Bunny-2BLLM=Qwen-1.5-1.8B, #Sample=2.7M, #Param=∼2B2026.0158.5
LLaVA-1.5-7BLLM=Vicuna-1.5-7B, #Sample=1.2M, #Param=≥7B2026.0158.3
MoE-LLaVA-2BLLM=Qwen-1.5-1.8B, #Sample=2.2M, #Param=∼2B2026.0157.3
Qwen-VL-ChatLLM=Qwen-7B, #Sample=1450M, #Param=≥7B2026.0156.7
CogVLMLLM=Vicuna-7B, #Sample=1500M, #Param=≥7B2026.0153.8
VILA-3BLLM=LLaMA-2.7B, #Sample=51M, #Param=∼3B2026.0152.7
Mini-Gemini-2BLLM=Gemma-2B, #Sample=2.7M, #Param=∼2B2026.0151.3
LLaVA-FA-1BLLM=Qwen-2.5-3B, #Sample=5M, #Param=∼1B2026.0149.4
Imp-3BLLM=Phi-2-2.7B, #Sample=1.6M, #Param=∼3B2026.0146.7
MoE-LLaVA-3BLLM=Phi-2-2.7B, #Sample=2.2M, #Param=∼3B2026.0141.8
SPHINX-TinyLLM=TLLaMA-1.1B, #Sample=15M, #Param=∼1B2026.0137.8
Bunny-3BLLM=Phi-2-2.7B, #Sample=2.7M, #Param=∼3B2026.0137.2