CLIP (Contrastive Language-Image Pretraining)

[klip] /klɪp/

Released in 2021 by OpenAI, CLIP jointly encodes natural language descriptions and images into a common latent space to make it easier to search for, classify and describe images. Source: Alban Leveau-Vallier (2025) ‘Glossary,’ trans. Michelle Noteboom, 'The World Through AI', JBE Books