List your company
ResearchFormerly WizWand

Multimodal Understanding on MME-RealWorld English

59.9AccuracyMiMo-VL-Miloco-7B
27.722436.076244.4352.7838Aug 5, 2025Aug 27, 2025Sep 19, 2025Oct 12, 2025Nov 3, 2025Nov 26, 2025Dec 19, 2025
Updated 7mo ago

Evaluation results

MethodLinks
MiMo-VL-Miloco-7BChain-of-thought reasoning=false2025.1259.9—
MiMo-VL-7B-SFT-2508Chain-of-thought reasoning=false2025.1259.7—
Full PrecModel Backbone=Qwen2.5-VL-7B-Instruct, Quantization Method=Full Precision, Quantization Bit-width/Group-size=None, Zero-shot=true2025.0858.55—
VLMQModel Backbone=Qwen2.5-VL-7B-Instruct, Quantization Method=VLMQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0858.32—
GPTQModel Backbone=Qwen2.5-VL-7B-Instruct, Quantization Method=GPTQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0858.24—
Qwen2.5-VL-7BChain-of-thought reasoning=false2025.1257.4—
Full PrecModel Backbone=LLaVA-OneVision-7B, Quantization Method=Full Precision, Quantization Bit-width/Group-size=None, Zero-shot=true2025.0857.39—
VLMQModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=VLMQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0857.34—
Full PrecModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=Full Precision, Quantization Bit-width/Group-size=None, Zero-shot=true2025.0857.3—
MBQModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=MBQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0856.79—
AWQModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=AWQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0856.72—
GPTAQModel Backbone=Qwen2.5-VL-7B-Instruct, Quantization Method=GPTAQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0856.32—
InternVL3-8BChain-of-thought reasoning=false2025.1256.1—
GPTAQModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=GPTAQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0855.83—
GPTQModel Backbone=LLaVA-OneVision-7B, Quantization Method=GPTQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0855.2—
GPTAQModel Backbone=LLaVA-OneVision-7B, Quantization Method=GPTAQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0854.82—
GPTQModel Backbone=Qwen2-VL-7B-Instruct, Quantization Method=GPTQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0854.47—
VLMQModel Backbone=LLaVA-OneVision-7B, Quantization Method=VLMQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0854.26—
Gemma-3-27B-ITChain-of-thought reasoning=false2025.1251.9—
GPTAQModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=GPTAQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0844.12—
Full PrecModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=Full Precision, Quantization Bit-width/Group-size=None, Zero-shot=true2025.0840.27—
GPTQModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=GPTQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0837.8—
VLMQModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=VLMQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0837.75—
AWQModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=AWQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0830.56—
MBQModel Backbone=Qwen2-VL-2B-Instruct, Quantization Method=MBQ, Quantization Bit-width/Group-size=INT3g128, Zero-shot=true2025.0828.96—
Innovator-VLvariant=8B-Instruct, parameters=8B2026.01—63.7
Innovator-VLvariant=8B-Thinking, parameters=8B2026.01—63.04
Intern-S1variant=mini, parameters=9B2026.01—56.39
InternVL3.5parameters=8B2026.01—62.54
LLaVA-OVversion=1.5, parameters=8B2026.01—61.57
MiMo-VLtraining=7B-SFT, parameters=7B2026.01—57.47
MiMo-VLtraining=7B-RL, parameters=7B2026.01—59.55
MiniCPM-Vversion=4.5, parameters=8B2026.01—60.94
Qwen3-VLparameters=8B2026.01—63.29