Meta AI text input now lets users interleave images with text

ezyang · x · 2026-07-21

Meta AI’s text input now accepts images interleaved directly with text, which the author says feels much more natural than having images handled as separate attachments.

The post is mainly a product UX observation, but it highlights a meaningful multimodal interface change: users can compose one input stream instead of treating images as external add-ons.

Related event: Meta AI text input now lets users interleave images with text(2 posts)→

Original post →

More from Multimodal

Multimodal channel →