Vision-Language Model (VLM)
A vision-language model jointly processes images and text, letting it answer questions about pictures, read documents, describe scenes, and connect visual content with language. It typically pairs a vision encoder with a language model so visual features and text tokens are reasoned over together in one context.
Definition
Vision-Language Model (VLM)Vision-language models bridge computer vision and natural language processing. An image is passed through a vision encoder that produces features aligned to the language model's representation space, so the model can attend to image content just as it attends to text. This enables visual question answering, chart and document understanding, and grounded description.
Foundational work such as CLIP learned aligned image-text representations from large web data, and later VLMs integrate such encoders directly into generative language models. VLMs are a central case of multimodality and power features like analyzing screenshots, diagrams, and photographs.
Frequently asked questions
What is Vision-Language Model (VLM)?
A vision-language model jointly processes images and text, letting it answer questions about pictures, read documents, describe scenes, and connect visual content with language. It typically pairs a vision encoder with a language model so visual features and text tokens are reasoned over together in one context.
Sources
Related terms
← Full AI & prompt engineering glossary
Put Vision-Language Model (VLM) to work with Prompeteer, the Agentic Contextual AI Platform →