# Zero-shot image-text matching

MobileCLIP-S0 is an efficient image-text model achieved through multimodal reinforcement training, significantly improving speed and size efficiency while maintaining high performance.