Our new X account is live! Follow @wizwand_team for updates

Multi-modal Retrieval (Image Query) on MMQA

76.52Recall@20

CLIP (ViT-L/14)

Updated 2d ago

Evaluation Results

Method	Links
CLIP (ViT-L/14) 2025.11		76.52	58.88	63.05
Long-CLIP-L 2025.11		73.91	54.04	58.63
Long-CLIP-L 2025.11		71.74	49.38	54.52
Long-CLIP-B 2025.11		66.96	50.72	54.51
CLIP (ViT-B/32) 2025.11		66.09	45.03	49.86
Long-CLIP-B 2025.11		66.09	47.67	51.94
CLIP (ViT-L/14) 2025.11		62.17	43.32	47.7
CLIP (ViT-B/32) 2025.11		52.61	36.65	40.32
Cohere Embed 3 2025.11		52.17	27.24	32.92
Cohere Embed 3 2025.11		52.17	28.17	33.61
Cohere Embed 3 2025.11		50.43	20.79	27.61
BLIP 2025.11		43.48	27.45	31.15
E5-V 2025.11		41.74	21.33	25.91
E5-V 2025.11		41.74	21.55	26.09
BLIP 2025.11		39.57	23.97	27.54
E5-V 2025.11		38.7	17.34	22.06
CLIP (ViT-L/14) 2025.11		1.3	0.41	0.62
Long-CLIP-L 2025.11		0.43	0.11	0.19
CLIP (ViT-B/32) 2025.11		0	0	0
Long-CLIP-B 2025.11		0	0	0
BLIP 2025.11		0	0	0