List your company

Multimodal Understanding on ShopBench

73.8ShopFront ScoreSeed 1.8*
36.04845.84955.6565.451Jan 29, 2026

Evaluation results

MethodLinks
Seed 1.8*Architecture=MoE2026.0173.857.963.570.163.969
Ostrakon-VLParam=8B, Architecture=Dense2026.016549.159.149.653.360.1
Gemini3-Pro*Architecture=MoE2026.0164.770.653.465.862.363.6
GLM-4.6VParam=106B, Architecture=MoE2026.016451.157.35754.160
InternVL3.5-241BParam=241B, Architecture=MoE2026.0163.552.762.959.754.261.1
Qwen3-VL-235BParam=235B, Architecture=MoE2026.0163.250.358.653.153.359.4
GLM-4.6V-FlashXParam=9B, Architecture=Dense2026.0161.947.65756.547.157.9
Ovis2.5-9BParam=9B, Architecture=Dense2026.0161.847.656.457.349.857.9
Qwen2.5-VL-72BParam=72B, Architecture=Dense2026.0160.849.557.349.350.757.3
Qwen3-VL-8BParam=8B, Architecture=Dense2026.0159.346.954.146.250.755.3
InternVL3-78BParam=78B, Architecture=Dense2026.0158.35154.849.249.855.7
GPT-5*Architecture=MoE2026.0157.256.265.36948.358.8
Qwen3-VL-3BParam=3B, Architecture=Dense2026.0156.944.653.847.949.853.6
Gemini2.5-Pro*Architecture=MoE2026.0152.256.565.961.556.755.9
InternVL3.5-8BParam=8B, Architecture=Dense2026.015246.350.141.347.549.9
InternVL3.5-4BParam=4B, Architecture=Dense2026.015044.348.739.344.547.9
MiniCPM-V-4Param=4B, Architecture=Dense2026.0149.242.34436.943.646.2
LLava-OneVisionParam=7B, Architecture=Dense2026.0137.539.241.627.135.237.6