Vision-language model
A vision-language model (VLM) is a model that takes images and text as input and produces text, so it can describe pictures, answer questions about them, read documents and extract data from screenshots.
Also called: VLM
By Divya Sharma · Checked Nov 15, 2026 · How we review
VLMs are what let a chatbot read a photo of a receipt or a scanned contract. They are a core tool for document parsing, where they increasingly replace or support classic OCR.
How it works
A typical VLM has a vision encoder that turns an image into a set of vectors, a small connector that maps those vectors into the language model's input space, and a large language model that reads them alongside the text prompt. LLaVA, from 2023, is a well-known open example: it connects a vision encoder to an LLM and was trained on instruction-following data about images. DeepMind's Flamingo in 2022 described itself as a family of visual language models.
VLM vs OCR
OCR turns pixels into characters. A VLM can also understand layout, tables, charts and handwriting, and answer questions about them. Some VLMs are tuned for documents, such as DeepSeek-OCR 2.
Limits
VLMs can misread small text, mix up numbers in dense tables or describe details that are not in the image, a visual form of hallucination. Check extracted figures before relying on them. VLMs are one kind of multimodal model.
Related terms
- Multimodal model
- A multimodal model is an AI model that can take in, and sometimes produce, more than one kind of data, such as text, images, audio or video, within a single model rather than through separate systems.
- Optical character recognition
- Optical character recognition (OCR) is software that reads text from images, such as scanned pages, photos or screenshots, and turns it into machine-readable characters that can be searched, copied or fed to a model.
- Document parsing
- Document parsing converts files such as PDFs, scans, slides and spreadsheets into structured text that software can use, keeping headings, reading order, tables and images instead of a flat stream of characters.
- Hallucination (AI)
- A hallucination is output from an AI model that sounds fluent and confident but is false, unsupported by its sources or made up, such as an invented citation, quote, statistic or API function.
- Table extraction
- Table extraction is the task of finding tables in documents and rebuilding their rows, columns, headers and cells as structured data, such as HTML, Markdown or CSV, instead of loose lines of text.