These are reusable skeletons for Seedance, ByteDance's text/image/audio-to-video model. The structure is done — you fill in the blanks. Each one follows the same formula: Subject + Action + Setting + Camera + Lighting + Audio, then a (aspect, resolution, seconds) note at the end. Swap the [BRACKETS], set your aspect ratio and duration in the app, and render.
They're written for Seedance 2.0 (native audio+video in one pass, 3–10s clips) and Seedance 2.5 (single clips up to 30s with scene changes). For finished versions, see the 35 best Seedance prompts; for the quick reference, the Seedance prompt cheat sheet.
How to use these
Copy a skeleton, swap every [BRACKET] for your real subject, action, and setting, and keep the camera, lighting, and audio parts in the order they appear. Write spoken lines in quotes — lip-sync is automatic in 8+ languages. Prefer positive phrasing over negatives, and keep no subtitles, no on-screen text on anything with dialogue or graphics.
One reminder that catches everyone: the (16:9, 1080p, 8s) tail is a note, not a command. Seedance has no --ar flag — aspect ratio, resolution, and duration are set in Dreamina, fal.ai, or whatever app you're using. Match the parenthetical to your actual settings. Full walkthrough in the realistic-video guide.
Cinematic shot
Single-subject, single-move shots with film lighting and layered ambient sound. Best for hero moments, mood pieces, and trailer beats at 5–10 seconds.
1. Character mood shot
[SUBJECT with wardrobe details] [ACTION, one continuous motion] in [SETTING] at [TIME OF DAY]. [CAMERA FRAMING], [ONE CAMERA MOVE]. [LIGHTING STYLE], [LENS / DEPTH detail]. Ambient: [BACKGROUND SOUND]. SFX: [FOREGROUND SOUND]. No subtitles, no on-screen text. (16:9, 2K, 8s)Swap: [SUBJECT] is who's on screen; [ACTION] is one clear motion; [SETTING]/[TIME OF DAY] place it; [CAMERA MOVE] picks one dolly/pan/push; [LIGHTING] and [LENS] set the look; the two audio brackets layer background then foreground sound.
Filled-in example:
A lone figure in a wet trench coat walks slowly down a narrow neon-lit alley at night, hands in pockets, breath visible in the cold. Rain-slicked cobblestones reflect pink and cyan signage. Medium wide shot, slow dolly-in following behind the figure. Low-key lighting, practical neon, anamorphic flares, shallow depth of field. Ambient: steady rain, distant traffic, a dripping gutter. SFX: footsteps splashing through puddles. No subtitles, no on-screen text. (16:9, 2K, 8s)2. Sweeping establishing shot
Establishing shot of [PLACE / LANDSCAPE] with [KEY DETAIL in frame]. [CAMERA MOVE, e.g. slow aerial push-in] revealing [WHAT COMES INTO VIEW]. [LIGHTING / TIME OF DAY], [WEATHER / ATMOSPHERE], cinematic color grade. Ambient: [ENVIRONMENTAL SOUND]. Score: [MUSIC MOOD]. No subtitles, no on-screen text. (16:9, 1080p, 10s)Swap: [PLACE] and [KEY DETAIL] set the wide frame; [CAMERA MOVE] and [WHAT COMES INTO VIEW] give the reveal; [LIGHTING]/[WEATHER] set atmosphere; [ENVIRONMENTAL SOUND] and [MUSIC MOOD] fill the native audio track.
3. Slow-motion detail
Extreme close-up of [SUBJECT / OBJECT] as [ACTION happens in slow motion]. [CAMERA: macro, locked or slow drift]. [LIGHTING], high shutter clarity, shallow depth of field, [COLOR PALETTE]. Ambient: [QUIET ROOM TONE]. SFX: [SINGLE CRISP SOUND]. No subtitles, no on-screen text. (16:9, 2K, 5s)Swap: [SUBJECT/OBJECT] and [ACTION] are the moment you're isolating; [CAMERA] stays tight; [LIGHTING]/[COLOR PALETTE] set the mood; the audio brackets keep it near-silent with one sharp accent.
Product & ad
Clean, sellable shots that keep the product hero and readable. Best for launches, hero reels, and paid social. Add your brand voiceover in quotes or leave the audio to SFX and music.
4. Product hero spin
[PRODUCT with material/finish] sits on [SURFACE] against [BACKDROP]. Slow 360-degree orbit around the product, [SECOND MOVE, e.g. push-in on a detail]. [LIGHTING STYLE, e.g. soft studio softbox], crisp reflections, [COLOR ACCENT]. Ambient: quiet studio room tone. SFX: [SUBTLE PRODUCT SOUND]. No subtitles, no on-screen text. (1:1, 1080p, 8s)Swap: [PRODUCT] plus [SURFACE]/[BACKDROP] stage it; the two moves give the orbit-then-detail beat; [LIGHTING]/[COLOR ACCENT] set the studio look; [PRODUCT SOUND] adds a small tactile cue.
Filled-in example:
A brushed-aluminum wireless earbud case sits on a matte black stone slab against a deep charcoal backdrop. Slow 360-degree orbit around the case, then a smooth push-in as the lid clicks open to reveal the earbuds. Soft studio softbox lighting from the left, crisp edge reflections, a single warm amber accent light behind. Ambient: quiet studio room tone. SFX: a clean magnetic snap as the lid opens. No subtitles, no on-screen text. (1:1, 1080p, 8s)5. Product in use
[PERSON / HANDS] using [PRODUCT] to [DO THE TASK] in [REAL SETTING]. [CAMERA: over-the-shoulder or handheld follow], staying on the product. [NATURAL LIGHTING], authentic and warm, shallow depth of field. Ambient: [ROOM / LOCATION SOUND]. SFX: [ACTION SOUND of the product working]. No subtitles, no on-screen text. (9:16, 1080p, 8s)Swap: [PERSON/HANDS] and [PRODUCT] show the use; [DO THE TASK] and [REAL SETTING] make it believable; [CAMERA] stays close; the audio brackets sell the moment with location tone plus the product's own sound.
6. Ad with voiceover hook
[PRODUCT / SCENE] as [ACTION unfolds]. [CAMERA MOVE that builds momentum]. [LIGHTING], vivid brand colors, energetic pacing. Voiceover (confident, [TONE]): "[HOOK LINE]" then "[BENEFIT LINE]". Ambient: [LIGHT BACKGROUND BED]. Music: [UPBEAT STYLE]. No subtitles, no on-screen text. (9:16, 1080p, 10s)Swap: [PRODUCT/SCENE] and [ACTION] carry the visual; [CAMERA MOVE] adds drive; the two quoted lines are your spoken hook and benefit — lip-sync and audio render together; [TONE] and [MUSIC] set the vibe.
Talking character / dialogue
Seedance renders speech, lip-sync, and sound in one pass. Put every spoken line in quotes and keep no subtitles, no on-screen text so captions don't appear. For short clips use one camera move.
7. Single speaker to camera
[CHARACTER with appearance details] [SITS / STANDS] in [SETTING], looking into the lens. Medium close-up, [SLOW PUSH-IN or locked]. [LIGHTING], natural skin tones, shallow depth of field. They say, in [LANGUAGE], [EMOTION]: "[LINE OF DIALOGUE]". Ambient: [ROOM TONE]. No subtitles, no on-screen text. (16:9, 1080p, 8s)Swap: [CHARACTER] and [SETTING] frame the speaker; [LIGHTING] sets mood; [LANGUAGE] and [EMOTION] steer the delivery; the quoted [LINE] is what they say — lip-sync is automatic.
Filled-in example:
A woman in her thirties with short dark hair and a grey knit sweater sits at a sunlit kitchen table, looking into the lens. Medium close-up, slow push-in. Soft morning window light, natural skin tones, shallow depth of field. She says, in English, warm and reassuring: "I promise you, the hardest part is already behind us." Ambient: quiet kitchen room tone, a kettle beginning to warm. No subtitles, no on-screen text. (16:9, 1080p, 8s)8. Two-character exchange
[CHARACTER A] and [CHARACTER B] talk in [SETTING]. [CAMERA: over-the-shoulder, favoring the speaker]. [LIGHTING / MOOD]. A says, [EMOTION]: "[A'S LINE]". B replies, [EMOTION]: "[B'S LINE]". Ambient: [BACKGROUND SOUND of the place]. No subtitles, no on-screen text. (16:9, 1080p, 10s)Swap: [CHARACTER A]/[CHARACTER B] and [SETTING] stage the scene; [CAMERA] favors whoever speaks; the two quoted lines are the back-and-forth with an [EMOTION] each; [BACKGROUND SOUND] grounds it.
9. Narrator over action
[SUBJECT] [ACTION] in [SETTING], while an unseen narrator speaks. [CAMERA MOVE that follows the action]. [LIGHTING], [STYLE]. Voiceover (calm, [TONE]): "[NARRATION LINE]". Ambient: [ENVIRONMENTAL SOUND]. Music: [UNDERSCORE MOOD]. No subtitles, no on-screen text. (16:9, 1080p, 8s)Swap: [SUBJECT]/[ACTION]/[SETTING] carry the picture; [CAMERA MOVE] follows it; the quoted [NARRATION LINE] plays over the top with [TONE]; [MUSIC] sets the bed under the voice.
B-roll & nature
No people, no dialogue — texture, motion, and atmosphere you can cut anywhere. Best for backdrops, transitions, and ambient loops. Lean on ambient sound instead of speech.
10. Nature texture loop
[NATURAL SUBJECT, e.g. tall grass / waves / clouds] [MOVING under NATURAL FORCE, e.g. wind / tide]. [CAMERA: slow drift or locked wide]. [TIME OF DAY / LIGHTING], [COLOR PALETTE], cinematic depth. Ambient: [NATURE SOUND]. No dialogue, no subtitles, no on-screen text. (16:9, 2K, 10s)Swap: [NATURAL SUBJECT] and [NATURAL FORCE] give the motion; [CAMERA] stays gentle; [LIGHTING]/[COLOR PALETTE] set the look; [NATURE SOUND] is the whole audio track.
11. Urban B-roll
[CITY SCENE, e.g. crosswalk / rooftop / market] with [MOVEMENT in frame, e.g. passing traffic]. [CAMERA MOVE, e.g. slow parallax pan]. [LIGHTING / TIME], [WEATHER], moody color grade. Ambient: [CITY SOUNDSCAPE]. No dialogue, no subtitles, no on-screen text. (16:9, 1080p, 8s)Swap: [CITY SCENE] and [MOVEMENT] set the frame; [CAMERA MOVE] adds parallax; [LIGHTING]/[WEATHER] set atmosphere; [CITY SOUNDSCAPE] fills the ambient bed.
12. Abstract / macro texture
Macro shot of [TEXTURE / MATERIAL, e.g. ink in water / frost forming] as it [SLOWLY TRANSFORMS]. [CAMERA: locked macro or slow zoom]. [LIGHTING], [COLOR PALETTE], high detail, shallow depth of field. Ambient: [SUBTLE ROOM TONE]. SFX: [SOFT MATERIAL SOUND]. No dialogue, no subtitles, no on-screen text. (16:9, 2K, 5s)Swap: [TEXTURE/MATERIAL] and [SLOWLY TRANSFORMS] are the abstract motion; [CAMERA] stays macro; [LIGHTING]/[COLOR PALETTE] set the palette; the audio brackets keep it quiet with one soft accent.
Vertical social 9:16
Built for TikTok, Reels, and Shorts — vertical framing, a strong first frame, and a fast hook. Best for scroll-stopping opens. See the full pack in Seedance prompts for social media.
13. Scroll-stopping hook
[SUBJECT] [DOES SOMETHING SURPRISING] in [SETTING], starting mid-action for an instant hook. Vertical framing, [FAST CAMERA MOVE, e.g. whip-pan into a push-in]. [BRIGHT / HIGH-CONTRAST LIGHTING], punchy colors. Ambient: [ENERGETIC BACKGROUND]. SFX: [SHARP ACCENT on the action]. No subtitles, no on-screen text. (9:16, 1080p, 5s)Swap: [SUBJECT] and [DOES SOMETHING SURPRISING] are the hook; [SETTING] places it; [FAST CAMERA MOVE] adds energy; the audio brackets punch the first second.
14. Creator talking-head (vertical)
[CREATOR with appearance] talks directly to camera in [SETTING], upper body in frame. Vertical medium shot, slight handheld energy, subtle push-in. [BRIGHT EVEN LIGHTING], clean background. They say, [ENERGY]: "[HOOK LINE]". Ambient: [LIGHT ROOM TONE]. No subtitles, no on-screen text. (9:16, 1080p, 8s)Swap: [CREATOR] and [SETTING] set the frame; the quoted [HOOK LINE] is the opener with an [ENERGY]; keep it vertical and bright for social.
15. Aesthetic day-in-the-life
[FIRST-PERSON or over-shoulder view] of [EVERYDAY MOMENT, e.g. making coffee] in [COZY SETTING]. Vertical framing, [GENTLE HANDHELD MOVE]. [WARM NATURAL LIGHTING], soft film-like color, shallow depth of field. Ambient: [COZY SOUNDSCAPE]. Music: [LO-FI / CALM STYLE]. No subtitles, no on-screen text. (9:16, 1080p, 8s)Swap: [EVERYDAY MOMENT] and [COZY SETTING] set the vibe; [GENTLE HANDHELD MOVE] keeps it human; [COZY SOUNDSCAPE] and [MUSIC] carry the calm aesthetic.
Multi-shot with Seedance 2.5
Seedance 2.5 renders a single clip up to 30 seconds with scene changes and no stitching. Use time-coded segments — 0-5s: … 5-12s: … — to sequence shots. Each segment gets its own framing and move.
16. Three-beat story
A [SHORT STORY PREMISE] told in three shots. 0-5s: wide establishing shot of [SETTING], [CAMERA MOVE], [LIGHTING]. 5-12s: medium shot of [SUBJECT / ACTION], [CAMERA MOVE], the moment builds. 12-18s: close-up on [PAYOFF DETAIL / EMOTION], [CAMERA MOVE], scene resolves. Consistent [COLOR GRADE] and [STYLE] across all shots. Ambient shifts with each scene: [S1 SOUND], [S2 SOUND], [S3 SOUND]. Music builds throughout. No subtitles, no on-screen text. (16:9, 1080p, 18s)Swap: [STORY PREMISE] is the arc; each time-coded segment gets its own [SETTING]/[SUBJECT]/[PAYOFF] plus a [CAMERA MOVE]; keep [COLOR GRADE] and [STYLE] consistent so the shots feel like one clip; the three ambient brackets shift per scene.
17. Product story sequence
A [PRODUCT] ad in four beats. 0-6s: [PROBLEM MOMENT] in [SETTING], [CAMERA MOVE], [MOOD LIGHTING]. 6-14s: [PRODUCT] appears and [SOLVES IT], smooth [CAMERA MOVE], brighter tone. 14-22s: [PRODUCT IN JOYFUL USE], energetic [CAMERA MOVE]. 22-28s: hero shot of [PRODUCT] with [BRAND COLOR] backdrop, slow orbit. Voiceover across the clip (confident, [TONE]): "[OPENING LINE]" … "[CLOSING TAGLINE]". Consistent brand color grade. Music builds from calm to upbeat. No subtitles, no on-screen text. (9:16, 1080p, 28s)Swap: [PRODUCT] runs through all four beats; each segment names a moment and a [CAMERA MOVE]; the quoted voiceover lines open and close the arc; keep one brand color grade so the 28-second clip reads as a single ad.
Reference-driven with @Image / @Video / @Audio
Seedance is multimodal, so you can attach references and mention them by name. @Image1 = character face/appearance, @Image2 = style/aesthetic, @Video1 = camera-movement or motion reference, @Audio1 = music/rhythm or dialogue track. Upload the matching files in the app.
18. Consistent character across shots
Use @Image1 as the character's face and appearance, and @Image2 for the overall style and color aesthetic. [CHARACTER] [ACTION] in [SETTING]. [CAMERA FRAMING], [ONE CAMERA MOVE]. [LIGHTING] consistent with the reference style. Ambient: [SOUND]. They say, [EMOTION]: "[LINE]". No subtitles, no on-screen text. (16:9, 1080p, 8s)Swap: @Image1 locks the face so the same character appears across clips; @Image2 sets the look; [CHARACTER]/[ACTION]/[SETTING] describe the shot; the quoted [LINE] is optional dialogue.
19. Match a motion and a track
Match the camera movement and pacing of @Video1, and sync the edit to the rhythm of @Audio1. [SUBJECT] [ACTION] in [SETTING]. [FRAMING], following the reference motion. [LIGHTING], [STYLE]. Ambient: [SOUND under the music]. No subtitles, no on-screen text. (9:16, 1080p, 10s)Swap: @Video1 supplies the camera move and pacing; @Audio1 supplies the music/rhythm to cut to; [SUBJECT]/[ACTION]/[SETTING] describe your new scene while the references drive motion and tempo.
Save the skeletons you reuse as a note or a preset in your video app so you only fill the changing brackets each time. For finished, ready-to-render prompts, browse the 35 best Seedance prompts, or keep the cheat sheet open while you build.
Frequently Asked Questions
How do I use these Seedance prompt templates?
Copy a skeleton, replace every [BRACKETED] placeholder with your real subject, action, and setting, then keep the camera, lighting, and audio parts in order. Each one follows Subject + Action + Setting + Camera + Lighting + Audio and ends with a (aspect, resolution, seconds) note. Delete any line you don't need, but keep at least one clear camera move and one audio cue.
Why does each template end with (16:9, 1080p, 8s) instead of a --ar flag?
Seedance has no inline parameters like Midjourney's --ar. Aspect ratio, resolution, and duration are set in the app or API — Dreamina, fal.ai, and other platforms — not in the prompt text. The parenthetical at the end is a reminder to match those settings; it doesn't control the render on its own, so set 9:16, 1080p, and your clip length in the interface too.
How do the @Image, @Video, and @Audio references work?
Seedance is multimodal, so you can attach references and mention them in the prompt. @Image1 locks a character's face or appearance, @Image2 supplies style or aesthetic, @Video1 is a camera-movement or motion reference, and @Audio1 supplies music, rhythm, or a dialogue track. Write plainly, for example "Use @Image1 as the character's appearance," and upload the matching files in the app.
Can I really get dialogue and sound from one prompt?
Yes. Seedance 2.0 was the first model to generate native audio and video in a single pass, so dialogue, sound effects, ambient sound, and music come out together with the picture — not added later. Put spoken lines in quotes and lip-sync is automatic across 8+ languages. Add "no subtitles, no on-screen text" so captions don't bleed onto the frame.
What's the difference between the single-shot and multi-shot templates?
Single-shot templates keep one camera move and suit 3s to 10s clips on Seedance 2.0. Multi-shot templates use time-coded segments like "0-3s: wide… 3-6s: medium… 6-10s: close-up" to sequence several shots from one prompt. Seedance 2.5 can render a single clip up to 30 seconds with scene changes and tempo shifts, so use the time-coded skeletons there.
How long can a Seedance clip be?
Seedance 2.0 renders clips at 3s, 5s, 8s, or 10s. Seedance 2.5, announced June 23, 2026, generates a single clip up to 30 seconds with scene changes and no stitching. Resolution runs 480p, 720p, or 1080p up to 2K, at 24 fps, in 16:9, 9:16, 1:1, or adaptive.
Should I use negatives like "no blur" in these templates?
Prefer positive phrasing — describe what you want in frame rather than listing what to avoid. Say "sharp focus, steady handheld" instead of "not blurry." The one negative worth keeping is "no subtitles, no on-screen text" on any clip with dialogue or graphics, because that reliably stops captions from appearing over the picture.