Skip to content

Aya Vision

A multilingual vision model covering 23 languages

Visit official site

What Aya Vision is

The model is described as a multilingual multimodal vision model from the Cohere For AI team, released to support global research.

What you can do with it

  • Ask a visual question about an image
  • Write a caption in another language
  • Start from the open weights for research

Pros & cons

✓ What we like

  • Efficient compute even with limited resources
  • Image captioning and visual question answering are named

! What to watch out for

  • It is focused on multilingual multimodal tasks
  • Versions named in setup are 8B and 32B

FAQ

How many languages are covered?

The material states 23 languages.

Which versions can be picked?

The setup steps name an 8B and a 32B version.

Last reviewed: 2026-11-15

More AI image generation tools

View all →

How we review