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

Multimodal Capability Evaluation on MM-Star

60.6Average Score

Qwen2.5-VL-7B

Updated 1mo ago

Evaluation Results

Method	Links
Qwen2.5-VL-7B 2026.03		60.6
Qwen2.5-VL-7B + SFT 2026.03		60.4
Qwen2.5-VL-7B + SFT+DPO+GRPO 2026.03		60.27
Qwen2.5-VL-7B + SFT+DPO 2026.03		59.73
Qwen2.5-VL-7B + SFT+GRPO 2026.03		59.33
Qwen2.5-VL-7B + SPA-VL 2026.03		59.13
Qwen2.5-VL-7B + SaFeR-VLM 2026.03		58.87
Qwen2.5-VL-3B + VLGuard 2026.03		52.67
Qwen2.5-VL-7B + TIS 2026.03		52.4
LLaVA-NeXT 2024.05		51.6
Qwen2.5-VL-3B + SFT 2026.03		51.33
Qwen2.5-VL-3B + SFT+GRPO 2026.03		51.07
Qwen2.5-VL-3B + SFT+DPO+GRPO 2026.03		51
Qwen2.5-VL-3B + SFT+DPO 2026.03		50.87
GPT-4V 2024.05		50.4
Qwen2.5-VL-7B + VLGuard 2026.03		49.87
Qwen2.5-VL-3B 2026.03		49.4
Qwen2.5-VL-3B + SPA-VL 2026.03		47.2
MiniGemini 2024.05		45.5
Qwen2.5-VL-3B + SaFeR-VLM 2026.03		44.13
Qwen2.5-VL-3B + TIS 2026.03		42.33
OmniLMM + RLAIF-V 2024.05		40.9
OmniLMM 2024.05		39.7
LLaVA 1.5 + RLAIF-V 2024.05		35.4
AMP-MEG 2024.05		34.8
Qwen-VL-Chat 2024.05		34.5
POVID 2024.05		34.3
LLaVA-RLHF 2024.05		34.2
VCD 2024.05		33.8
Silkie 2024.05		33.6
LLaVA 1.5 2024.05		33.3
RLHF-V 2024.05		33.2
Less-is-more 2024.05		32.9
OPERA 2024.05		32.9
HA-DPO 2024.05		32.9
CCA-LLaVA 2024.05		32.1