Picture it — receive the guide
The image is already
in your head.
You can’t describe a picture — nobody can. But you can describe a world: who’s there, where, when, what’s happening. Tell it as a story, in any medium you imagine, and receive a complete production guide for the exact image you already see.
One idea · any medium
Your idea, rendered as
one world · one look · one cast
Overview
One world, many frames — in any medium
Not a photo tool. A whatever-you-see tool.
The genie reads the medium straight from your words and locks it into every frame — then holds the cast, the light and the look identical across the whole set. Say nothing and it picks the medium your idea deserves.
The idea
You think in scenes, not in prompts
Describe a world, not a picture.
Ask someone to describe a picture and they freeze — a frozen frame has no handholds. Ask what’s happening — who’s there, the hour, what just moved — and it pours out. The mind holds imagery as situations, not flat stills.
Research on mental imagery agrees: narrate a scene as a small story and you recruit a far richer, more coherent picture in your head — and everything you hand a model inherits that coherence. A described world lands markedly closer to what you actually saw than a pile of keywords ever will.
01 — A picture
“a knight, castle, dramatic, 8k, cinematic”
Keywords, no world. The model guesses — differently every time.
02 — A world
“dawn after a siege; one exhausted knight at a broken gate, mist off the moat”
A situation. Coherent, faithful, repeatable across a whole set.
03 — The render
same story · different shots
One world, one look, one cast — many frames that belong together.
The method
From a sentence to a shoot
Tell it
Say what’s happening, to whom, and in what style — photo, painting, 3D, anime, surreal, anything. Add a reference if you have the look.
See your world
The genie writes it back as a one-page treatment — the medium, the world, the cast, the look, the palette, the rules that hold it together.
Correct it, in plain words
Read it and reply like you’re talking to a director: “colder, drop the third figure.” No sliders, no prompt syntax.
Shoot the frames
Pick how many frames. It builds a shot list, then a full, model-native guide for every frame — one world, one medium, one look.
The apparatus
Built to be believed
Same story, different shots
One anchor frame sets the cast and the look; every other frame inherits it — a set that belongs together, not six unrelated images.
A treatment you can edit
Your world returns as a readable page. Change the medium, the light, the cast — in plain language, before a single frame is shot.
Model-native guides
Each guide is written in your image model’s own dialect — a right-sized prompt, palette with hex, the failure modes and the exact fixes.
Bring your own reference
Upload the look and the genie reads it at full fidelity — its exact style, colour, grain and light — then tells you precisely when to attach it.
Consistency, handled
One identical style tail runs through the set, and the anchor keeps faces, wardrobe and world locked frame to frame.
It waits for you
Refresh, close the tab, come back next week — your world is saved. Shoot more frames in it whenever the idea returns.
Your turn
Say it as a story.
One sentence is enough to begin. The genie takes it from there.
Enter the studio