Describe subject, a second-by-second action timeline, camera, environment, and audio in one block, with negative constraints, so the shot is controllable instead of a lucky roll.
The prompt
Generate one continuous 8-second cinematic shot. Subject: [SUBJECT]. Action timeline: 0-2s [ACTION], 2-6s [ACTION], 6-8s [ACTION]. Camera: [LENS + MOVEMENT]. Environment: [PLACE + WEATHER]. Audio: [AMBIENCE]. Keep identity, wardrobe, and lighting consistent. No cuts, subtitles, watermarks, or extra limbs.
Replace the variables
[SUBJECT]The main subject.
[LENS + MOVEMENT]e.g. 35mm slow dolly-in.
Which model to use
- Veo 3.1 — Video generation is a Veo capability, not a text-model one.
Example input
- [SUBJECT] = "a barista finishing a latte". Camera = "slow push-in".
Example output
We only publish example outputs from real model runs. This template has not been formally tested yet, so no output is shown. Run it yourself with the input above.
Why this structure works
- A timeline gives the model temporal control instead of one static idea.
- Negative constraints suppress the usual video artefacts.
Common failures & fixes
Motion is chaotic.
Fix: Simplify to one clear action per time band.
Known limitations
- Video output is non-deterministic; expect several tries per usable shot.