Multimodal
A model that handles more than one kind of input or output: text, image, audio, video.
A multimodal model accepts or produces more than one modality. Most current examples accept text and images, and some accept audio as well. A vision language model encodes an image into the same token space as text, so the transformer can reason across both.
The practical value is removing a translation step. Instead of describing a screenshot in words and hoping the model reconstructs it, you hand over the screenshot. That alone changes what is possible for document work, interface review, and any task where visual detail matters.
The cost is usually tokens: images and audio consume far more of the context window than the equivalent text, and image-heavy workflows are often the most expensive thing you run.
Related terms
LLM
A model trained on a huge amount of text to predict the next token. The engine behind most AI products.
Vision Model
A model that reads images, turning them into something a language model can reason over.
Context Window
Everything the model sees on one request. Finite, and it decides what fits.
Cost per Token
What you pay for each unit of text the model reads or writes.
Want this working inside your team?
It starts with one conversation about your workflow.