Auto color-grade your portraits with a cinematic prompt.
Upload a character photo and an environment photo, dial in the grade, mood, depth of field and more, and get a ready-to-paste film-still prompt. The AI reads both photos to tailor the wardrobe, placement, and in-scene light — you keep the cinematic craft.
- No sign-up
- Photos never stored
- 100% free
Your prompt appears here
Add both photos, pick a look, and generate. You can edit the prompt right here before copying it.
- Add a character and an environment photo
- Dial in the look (or load a preset)
- Generate, then copy into your image tool
From two photos to a film still
Add two photos
A character portrait to preserve, and an environment to use as the setting.
Dial in the look
Pick the grade, mood, depth of field, shot size, falloff and more from the menus.
Copy the prompt
The AI reads both photos, fills the template, and hands you a prompt to paste.
What happens to your photos
Short version: nothing sticks around.
- Never saved. Both photos are processed in memory and discarded with the request — no uploads folder, no database.
- Metadata stripped. Photos are re-encoded first, which removes location and camera data.
- Read by Google’s Gemini AI. That’s what tailors the wardrobe, placement and lighting lines to your references.
Frequently asked questions
What does the cinematic prompt generator do?
It builds a ready-to-paste text-to-image prompt for a single cinematic film still. You give it a character photo and an environment photo, pick the look (grade, mood, depth of field, shot size, and more), and it writes a detailed prompt that preserves the person's identity, treats the location as the setting, and applies a deliberate color grade — the kind of prompt that takes real craft to write by hand.
How are my photos used?
They're re-encoded in memory (which strips location and camera metadata), sent to Google's Gemini AI to read once and tailor the wardrobe, placement, and in-scene lighting lines to your references, then discarded with the request. Nothing is saved to our servers or shown to anyone.
Which image is the character and which is the environment?
There are two clearly labelled upload fields. The environment photo becomes the setting (its architecture and depth), and the character photo's face and identity are preserved. In the generated prompt they map to "image 1 = environment" and "image 2 = character", so paste your images in that order into your image tool.
Where do I paste the finished prompt?
Into any capable image model that accepts reference images — for example an image-generation tool that lets you attach two reference photos. Attach the environment as the first image and the character as the second, then paste the generated prompt.
The face keeps changing when I generate the image. What do I do?
A heavy grade plus a face from a separate reference can make identity drift. Run it in two passes: first lock the likeness and composition, then a second pass that only applies the grade and focus treatment while keeping the face identical. The tip appears with every prompt you generate here.
Do I need to pay or sign up?
No. The cinematic prompt generator is completely free, with no account and no email required.