Generated images are great — but sometimes the right visual already exists: your hero photography, product shots, campaign stills. When an agent needs an image now, it asks which you'd prefer, and the image picker opens your brand's library right inside the chat.
The whole flow in ~20 seconds: ask for a post, choose your image source, pick from the library, and the design builds from your photo.
1. The agent asks before it assumes
Ask for anything visual — a social post, a mockup — without specifying an image, and the agent puts the choice to you:
Use an existing brand image — continuity and credibility from photography you already own.
Create a new image — the agent generates a fresh on-brand visual instead.
2. Your library, in the chat
Choose the existing-image route and the picker renders inline — the image library from your Brand Hub, scrollable, with the full collection one tap away:
Click an image and the agent confirms it's building your design from that exact file — no uploads, no file paths, no leaving the conversation.
3. The design comes back built around your image
Whatever you were making — here, an Instagram post — comes back with your chosen photo as the visual foundation:
Tips for using it well
Stock the library. The picker is only as good as what's in it — upload hero photography, product shots and campaign stills to your Brand Hub and every visual task starts from genuinely on-brand material.
Mix both modes. Library images for credibility (real products, real people), generated images for concepts and campaign ideas that don't exist yet.
It works mid-flow too. The picker isn't just for the first ask — agents offer it whenever a visual decision comes up inside a longer piece of work.
Drafting social posts with these visuals? The companion guide covers the full publish flow: Draft and publish social posts directly from chat.