Gallery Overview

How the community gallery works, what metadata is shown, and how privacy controls map to the product's sharing settings.

Gallery structure

The gallery displays publicly shared images in a card grid. Each card shows:

Image preview — the generated result.
Prompt summary — a short text preview (expandable for the full prompt).
Generation metadata — model, ratio, style tags.
Author — the account that shared it (or "Anonymous" if the user chose to hide their name).
Date — when it was generated.

Sorting and layout options

Newest first — default feed order.
Most liked — sorted by the community heart / reaction count.
Grid / List — toggle between compact card view and a wider list with more prompt text visible.

What the "Private" and "Flagged" labels really mean

The product does not have a gallery-native private/flag workflow. Privacy is a generation-level setting controlled before you submit:

Shared (default on free tier) → image appears in the public gallery after generation.
Private → image never appears in the gallery; it lives only in My Works. Paid plans can lock sharing off by default in account settings.

There is no per-image "flag for review" toggle in the UI. If a community image violates policy, use the Report button on the card to flag it for moderation.

Card actions

Button	What it does
Use prompt	Copies the prompt (and parameters if supported) into the editor — the fastest remix workflow.
Heart / Like	Adds a reaction; contributes to "most liked" sort.
Save / Favorite	Adds to your Favorites tab in My Works (not a gallery state — a personal collection).
Share	Generates a shareable link for the card.
Report	Flags the image for moderation review.

Gallery vs. My Works — where images live

Google's Gemini image generation guide describes a chat-based, multi-turn workflow: you generate, then feed the output back as the next input for an edit. That workflow lives in the editor, not the gallery.

Gallery = discover + borrow prompts from others; images are public by default.
My Works = your private generation history; all images default to private until you explicitly share them.

Iterating on gallery images (Gemini-style multi-turn)

The gallery is a prompt library, not an editing surface. To iterate on a gallery card the way Google's examples show:

Click Use prompt — the prompt and parameters load into the editor.
Generate your own version.
Optionally attach the gallery image as a reference alongside the prompt (up to 14 references in the editor).
Iterate with text: "same scene, but warmer light" or "add a blue scarf to the subject."

This mirrors Google's chat + multi-turn editing pattern without leaving the product.

Search and attribution

Use Search and filter to find images by style, model, or keyword.
Gallery prompts are public. If you remix a prompt that is clearly brand-owned or professional work, note the source in your internal briefs — see Copy and use prompts for attribution guidance.