List your company

Multimodal Multi-choice on MMStar (Accuracy)

65.1AccuracyClaude3.7-Sonnet
45.75650.77855.860.822Dec 18, 2025

Evaluation results

MethodLinks
Claude3.7-Sonnet2025.1265.1
SkiLa2025.1264.8
SkiLa-V2025.1264.5
LLaVA-OneVision 7BModel Size=7B2025.1261.9
GPT-4o2025.1261.6
Direct SFT2025.1261.4
LVR 7B*Model Size=7B, Tested by authors=true2025.1261.3
Vision-R1 7B*Model Size=7B, Tested by authors=true2025.1260.7
Qwen2.5-VL 7BModel Size=7B2025.1260.3
Gemma3 27BModel Size=27B2025.1259.6
GPT-4v2025.1256
GPT-4o-mini2025.1254.8
ROSS 7BModel Size=7B2025.1253.9
Cambrian 13BModel Size=13B2025.1247.1
Janus-Pro 7BModel Size=7B2025.1246.5