01 / The Model
Multimodal
Able to work with more than text: images, audio, scanned documents, sometimes video. You can hand it a photo of a whiteboard or an invoice and ask about it.
How you’ll hear it
“Can it read our scanned forms?”
“It is multimodal, so yes. Check the numbers it pulls out.”
Avoid
Assuming every model can read images or audio.
The longer version
A multimodal model accepts, and sometimes produces, several kinds of input. That opens up reading scanned paperwork, describing photos and transcribing meetings. Accuracy varies a lot by format and quality, so test it on your own material.