Launching today
Cinemuse by DIGITOLGA
Save any image, get its prompt, recreate it
2 followers
Save any image, get its prompt, recreate it
2 followers
Right-click any image on the web to save it to a board. Seconds later Cinemuse writes a 10-part prompt for it, from camera angle to lighting, palette and lens. Hit Recreate to regenerate the frame with up to 11 image models using your own API keys. Free, no account, 17 languages.
Free
Launch Team
AWS Startup AdvisorPersonalized AWS guidance that meets you where you’re already building
Promoted
Hi everyone, I'm Olga, an AI film director.
My references used to live in a folder called "New folder (3)": hundreds of frames I'd saved for the light, the lens or the colour, and no idea where half of them came from. Every time I wanted that look in a generation, I had to work the prompt out by hand.
So I built Cinemuse. Right-click an image on any site and it lands on a board. A few seconds later you get a prompt in ten parts (image type, camera angle, subject, clothing, pose, setting, colour tone with HEX codes, lighting and mood, effects, film stock and lens). Then you can recreate the frame with up to 11 image models and compare the takes next to the original.
The latest update came straight from the first users: you can drag a whole folder of references onto a board, crop black bars off film stills (the prompt rewrites itself for the new frame) and select cards with the mouse like in Finder.
Things I cared about while building it:
No account and no server. You plug in your own API keys and they stay in your browser.
You see the price before every generation.
The interface works in 17 languages.
It's free. I'd love to hear what you would add and which models you want next.