AI vision (image understanding)
Definition
AI vision is the ability of an AI model to look at an image you give it, such as a photo, screenshot, chart or scanned document, and describe it, read the text in it or answer questions about it.
By Ronan Liedmeier, fractional head of growth and course author · Published 17 October 2026 · Last updated 17 October 2026
Also called: image understanding, image input
At work it saves retyping: receipts, whiteboard photos and error screenshots can go straight into a chat or an automation. It is separate from image generation, which makes new images.
How it works
Models that accept images are multimodal. The image is converted into tokens that the model processes alongside your text, and you can send one image or several and ask the model to compare them. Combined with structured output, a vision model can pull fields such as supplier, date and amount from a photo into a spreadsheet row.
The vendors are open about the limits. Anthropic's documentation says Claude can make mistakes on low-quality, rotated or very small images, gives only approximate counts when there are many small objects, and should be checked by a person before high-stakes use.
Example
A property manager has caretakers photograph utility meters at the end of each month. An automation sends each photo to a vision model, which returns the meter ID and the reading. Photos the model marks as unclear, such as a reading shot through a dirty glass cover, go to the office for a manual check before the figures reach the tenants' bills.
Common mistake
Trusting every extraction because most are right. Glare, handwriting and low resolution produce confident misreads, and a 3 read as an 8 looks exactly like a correct value in a spreadsheet. Route unclear images to a person, and have someone spot-check a sample of the confident readings as well.
Learn it in AI Academy
- Working With ImagesBeginner · 10 min
- Voice & Image in WorkflowsAdvanced · 13 min
Related terms
- Multimodal AI
- Multimodal AI is AI that can work with more than one kind of input or output, such as text, images, audio and video, in the same conversation. You can show it a photo or screenshot and ask about it, or describe a picture and get one made.
- AI image generation
- AI image generation is the use of an AI model to create a new picture from a written description, or to edit an existing image by describing the change you want. You describe the subject, style and format in plain words, then refine the result in follow-up messages.
- Structured output (JSON mode)
- Structured output is a way of getting an AI model to return its answer in a fixed, machine-readable format, usually JSON with named fields, instead of free text. It lets another program read the answer reliably, for example to fill a spreadsheet row or pick the next step in a workflow.
- Human in the loop
- Human in the loop is a setup where an AI system pauses at set points so a person can approve or correct its work before it takes effect. It is used for actions that are hard to undo, such as sending emails to customers or moving money.