Text to prompt.
One sentence in, a full brief out.
Describe the shot in plain words — any language works. It becomes a complete specification: framing, subject, wardrobe, lighting, camera and style, ready to fine-tune and copy for any image or video model.
The builder and the prompts are in English — AI models follow English prompts best.
How text to prompt works.
Describe the shot
A sentence is enough: who, where, what light. Add as much or as little detail as you like.
Get a full specification
Claude expands it into around 90 concrete choices across framing, scene, subject, lighting, camera and style.
Fine-tune and copy
Change anything in the builder, then copy the prompt formatted for your model.
Start from a shot.
All examples →
Questions.
Can I write in my own language?+–
Yes. Describe the scene in any language; the finished prompt is written in English, which AI models follow best.
Does it work for video?+–
Yes. Pick a video model in the builder and the Motion section adds camera movement, action, duration and sound.
Is it free?+–
Yes. It uses your own Claude.ai chat, so there’s nothing to pay and no account to create.