Each entry is a complete, copy-ready master prompt — not a teaser. Open one, hit copy, paste it into your video model and change only the details you want to change.
∅ Nothing matches that yet. Try another keyword or reset the filters.
These prompts are written as complete systems. Three habits keep the output close to the intent.
Copy the entire prompt rather than a paragraph of it. Shortening a master prompt usually removes the part that was holding continuity together.
Swap the specific details — subject, location, wardrobe, duration — and leave the structure and camera language untouched.
[SUBJECT] — who or what is on screen[LOCATION] — one environment only[STYLE] — photoreal, 3D, documentary[DIALOGUE] — or explicitly “none”Most failed generations come from stacking changes mid-shot. Keep one location and one lighting rule for the whole clip, and give each beat a single camera behaviour.
If a model refuses a prompt, rerun the same text on a different model before rewriting it.
This is the complete prompt text, copied without edits.