Share your thoughts, 1 month free Claude Pro on usSee more

Multimodal Evaluation on MM-Vet v2

81.6Score

Vero Q3T-8B

Updated 19d ago

Evaluation Results

Method	Links
Vero Q3T-8B 2026.04		81.6
Vero Mi-7B 2026.04		78.7
Q3VL 8B-Thk 2026.04		74.5
GPT-5 Nano 2026.04		71.5
Vero Q3I-8B 2026.04		70.2
EVA 2026.05		69.3
EVA 2026.05		69.3
Vanilla 2026.05		69.1
Vanilla 2026.05		69.1
EVA(text) 2026.05		69
EVA(text) 2026.05		69
Q3VL 8B-Ins 2026.04		67.6
VLGuard 2026.05		67.5
VLGuard 2026.05		67.5
Vero Q25-7B 2026.04		66.6
JailGuard 2026.05		64.4
JailGuard 2026.05		64.4
Q25VL 7B-Ins 2026.04		62.9
MiMoVL 7B-RL 2026.04		61.2
Mo2-O 7B 2026.04		60.8
Vanilla 2026.05		56.9
Vanilla 2026.05		56.9
EVA(text) 2026.05		55.7
EVA(text) 2026.05		55.7
EVA 2026.05		55.5
EVA 2026.05		55.5
JailGuard 2026.05		54.5
JailGuard 2026.05		54.5
VLGuard 2026.05		46.6
VLGuard 2026.05		46.6
Adashield-A 2026.05		46.2
Adashield-A 2026.05		46.2
Adashield-A 2026.05		33.2
Adashield-A 2026.05		33.2
EVA(text) 2026.05		28.6
EVA(text) 2026.05		28.6
Vanilla 2026.05		28.5
Vanilla 2026.05		28.5
EVA 2026.05		28.3
EVA 2026.05		28.3
VLGuard 2026.05		28.1
VLGuard 2026.05		28.1
JailGuard 2026.05		26.5
JailGuard 2026.05		26.5
Adashield-A 2026.05		14.1
Adashield-A 2026.05		14.1