CLIP ViT-B/32

CLIP ViT-B/32 is OpenAI's vision-language embedding model that encodes images and text into a shared vector space for zero-shot image classification, image-text retrieval, and multimodal similarity tasks. It uses a Vision Transformer (ViT-B/32) image encoder paired with a text transformer encoder, trained via contrastive learning on 400 million image-caption pairs. Source: supabase.