What is Multimodal?

A model that understands more than text — images, audio, or documents — letting you feed it screenshots, photos, or scans alongside words.

How it works

Multimodal models turn unstructured input into structured understanding: read a receipt from a photo, describe a chart, extract data from a scanned document. It expands what an agent can perceive beyond plain text.

Example

In practiceYou photograph a paper invoice and the model extracts the line items, totals, and vendor into structured data.

When is multimodal worth it?

Any time your input isn't already clean text — receipts, screenshots, diagrams, handwriting, or scanned documents.

Go deeper: Deploy Your First AI Agent

The complete, step-by-step guide — every command, config, and the reasoning behind it.

Get the guide