LAION is a non-profit initiative that provides datasets, tools and models to foster artificial intelligence research. By promoting public education and the rational use of resources, LAION contributes to open and environmentally responsible research, relying on the reuse of existing datasets and models.
Explore a dataset of 400 million image-text pairs.
Analyze 5.85 billion multilingual image-text pairs.
Use the largest CLIP Vision Transformer model for your research.