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

Multimodal Reasoning Accuracy on MM-Vet

76.2Pass@1 Accuracy

InternVL3-9B

Updated 5d ago

Evaluation Results

Method	Links
InternVL3-9B 2026.04		76.2	-
Qwen2.5-VL-7B-Instruct + RFT 2026.04		73.8	-
Claude-3.5-Sonnet 2026.04		70.1	-
GPT-4o-20240513 2026.04		69.1	-
GPT-4V 2026.04		67.5	-
Qwen2.5-VL-7B-Instruct 2026.04		67.2	-
Qwen2.5-VL-7B-Instruct 2026.04		67.1	-
Qwen2.5-VL-7B-Instruct 2026.04		67.1	-
InternVL2.5-26B 2026.04		65	-
Qwen2.5-VL-7B-Instruct + cold start 2026.04		64.2	-
Gemini-1.5-Pro 2026.04		64	-
InternVL2.5-8B 2026.04		62.8	-
Qwen2-VL-7B 2026.04		62	-
LLaVA-OneVision-72B 2026.04		60.6	-
MiniCPM-V2.6 2026.04		60	-
Cambrian-34B 2026.04		53.2	-
mPLUG-Owl2 2024.12		-	36.2
LLaVA-1.5 2024.12		-	30.5
LLaVA-Phi 2024.12		-	28.9
Imp-v1 2024.12		-	33.1
MoE-LLaVA-3.6B 2024.12		-	35.9
TinyLLaVA 2024.12		-	32
Mipha-3B 2024.12		-	32.1
Olympus 2024.12		-	33.8