Give the editor a prompt and it writes the hooks, generates the images and lays out every frame with consistent typography. Then open any frame as a canvas and change whatever you want — text, layers, colors, backgrounds — without regenerating the rest.
A prompt produces a cover, the body slides and a closer, with the same font treatment and margins across all of them. Consistency is the part a general image model never gets right.
02
Edit without regenerating
Fix a typo, move a headline, swap a background, restyle the text — the other frames stay exactly as they were. No re-rolling a set of five images because one word was wrong.
03
Your brand, not a template
Pull from your own image packs and product catalog so the generated frames use your imagery, and keep captions in the language you picked per slideshow.
What you control
Every layer of a frame
The canvas is the same one the app runs on — what you see while editing is what renders when you publish.
Type
Text style
Weight, color, casing and alignment per frame
Text background
Solid, highlight or none, matched across slides
Captions
Written per platform, editable before publishing
Language
Generated in the language you pick, not translated after
Image
Backgrounds
Generated, uploaded, or pulled from an image pack
Models
Seedream, Nano Banana and Nano Banana Pro
Dimensions
Vertical, square and landscape output
Products
Feature a catalog item as the reference for a frame
Set
Frame count
As many slides as the story needs
Reordering
Move slides without breaking the styling
Batch caption edit
Rewrite copy across frames for 0.5 credits
Export
Straight to a scheduled post, or download the frames