PromptForge
AI Glossary

Models & architecture

Multimodal model

Also known as: multimodal, vision model, multi-modal

Definition

A multimodal model accepts or produces more than one type of data — commonly text plus images, and increasingly audio and video. It processes them in a shared representation rather than through separate pipelines.

Practically, multimodality means you can hand a model a screenshot, a chart, a PDF page or a photo and ask questions about it in natural language. This has quietly become one of the most useful capabilities in day-to-day work: debugging from a screenshot, extracting data from a photographed table, describing a diagram.

Capability varies a lot by modality and by task. Reading text in images (OCR-style tasks) is generally strong. Precise spatial reasoning — exact positions, counting many similar objects, reading cluttered charts — remains noticeably weaker than text reasoning.

Note that images consume tokens too, often substantially. A high-resolution image can cost as much as several pages of text.

Related terms

Put this into practice

Understanding the term is step one. Our free courses and tools let you actually use it.