Quick-Start Templates

Copy-paste templates for each input mode. Fill in the bracketed sections.

Text-to-Video (General)

[Subject: describe person/object with visual details]. [Action: concrete verb + intensity].
[Environment: location + lighting + atmosphere]. [Camera: one movement].
[Style: 2-3 aesthetic keywords]. Smooth motion, stable framing.

Image-to-Video

[What moves/changes in the scene]. [Camera: one movement or "camera holds fixed"].
Preserve composition and colors. [1-2 style words].

Reference Video Transfer

[New subject description] in [new environment]. Respect motion from reference:
reinterpret texture and color. [1-2 lighting words]. [1-2 style words].

Talking Head / Avatar

The person in [Video1] who looks like [Image1] speaks to camera.
Natural micro-expressions, subtle head movement, natural blinks.
[Background]. Camera holds fixed. [Lighting]. Avoid identity drift.

Timeline (10–15s)

[Global style: cinematic, 4K, film grain]
[0s–Xs]: [Shot type] — [camera]. [Scene.]
[Xs–Ys]: [Shot type] — [camera]. [Action.]
[Ys–Zs]: [Shot type] — [camera]. [Resolution.]

UGC Product Demo

0-4s dialogue: "[Hook sentence about the product]"
visuals: [subject holds/shows the product]. [Close-up]. [Natural lighting].

4-12s dialogue: "[Benefit explanation, casual spoken tone]"
visuals: [subject uses the product, camera pulls back to show context]. [Warm lighting].

12-15s dialogue: "[CTA or reaction]"
visuals: [subject faces camera, product visible]. Fixed framing.

Video Extension

Extend this video. Keep the same character, voice, and environment from the reference.
[New action or dialogue for the next section]. Smooth transition from the previous scene.

Parameters Reference

Parameter	Options	Default	Notes
`duration`	4–15s or -1 (auto)	5s	Use -1 to let model decide optimal length
`ratio`	16:9, 9:16, 1:1, 4:3, 3:4, 21:9, adaptive	16:9	9:16 for vertical/social content
`resolution`	720p, 480p	720p	Use 480p only for fast iteration
`generate_audio`	true/false	true	Set false if adding your own audio

Model Selection

Model	Use when
`dreamina-seedance-2-0-260128`	Quality matters — client work, hero videos, public-facing
`dreamina-seedance-2-0-fast-260128`	Rapid iteration, testing prompts, internal review

Recommended workflow: test with fast → refine → final render with 2-0.

Always verify the live model ID with pica model search "seedance" before generating.

Quick Start Templates