
Good Seedance 2.0 prompts read like a short shot list: one subject, one action, one place, one camera move, the light, the sound, and what to leave out. This guide gives you that formula and 16 copy-ready Seedance prompts, grouped by use.
This guide is independent and not affiliated with ByteDance. Three of the Seedance 2.0 prompts below come with real clips made in our studio. We wrote the other thirteen and have not published clips for them.
At a glance
- BytePlus recommends a prompt of no more than 1,000 English words or 500 Chinese characters. Our Seedance 2.0 prompts run 50 to 90 words, and these Seedance 2.0 prompts are free to copy.
- Seedance 2.0 makes 4 to 15 second clips with sound, and our studio offers 4, 8, 12 and 15 seconds.
- All 16 Seedance 2.0 prompts sit in quote blocks, one per paragraph, ready to paste.
- Draft on Seedance 2.0 Mini at 4 seconds and 480p, then render the keeper on Seedance 2.0.
- Change one thing per retake so you can see what helped.
What makes good Seedance 2.0 prompts?
Good Seedance 2.0 prompts name the same eight parts every time. ByteDance's launch post describes 15-second multi-shot output with background music, ambient sound and character voiceovers. That is why sound and shot order belong in the prompt, not only the picture.
The eight parts are:
- Subject: one person or object, with two or three visible details.
- Action: one verb phrase, such as "pours" or "turns toward the door".
- Scene: place and time of day.
- Camera: one move and one shot size.
- Light: source, color and mood.
- Style: "photorealistic", "hand-drawn anime", "documentary".
- Sound and speech: ambience, music, or a quoted line of dialogue.
- Exclusions: "no text, no subtitles, no logos".
The 1,000-word figure is BytePlus's recommended length, not a hard limit. Everything else here is general writing advice from our own practice, not an official rule. Shorter prompts are easier to debug.
What is a reusable formula for Seedance 2.0 prompts?
Fill this template for your own Seedance 2.0 prompts, then delete any slot you do not need:
[Style], [subject with 2 details] [one action] in [place, time]. [Camera move + shot size]. [Light]. [Sound / quoted dialogue]. [Exclusions].
For a multi-shot clip, number the shots and give each one a single job:
Shot 1: [framing, subject, action]. Shot 2: [framing, new angle or new action]. Shot 3: [payoff]. Same [subject, light, style] throughout. [Sound]. [Exclusions].
Keep the camera words plain in your Seedance prompts. These are common film terms you can reuse:
| Goal | Phrase to try | Good for |
|---|---|---|
| Build emotion | "slow push-in from wide two-shot to close-up" | Dialogue |
| Follow action | "tracking shot alongside at walking pace" | Fashion, travel |
| Show scale | "slow aerial pull-back" | Landscapes |
| Show a product | "slow 180-degree orbit, macro lens" | Ads |
| Stay calm | "locked-off static frame" | Talking scenes |
| Raw feel | "handheld, slight shake" | Street, chase |
Which Seedance 2.0 prompts work for dialogue and short drama?
Dialogue-focused Seedance 2.0 prompts need a quoted line, a named emotion and a slow camera. Seedance 2.0 generates audio with the picture, and Dreamina's feature page lists audio-driven timing for narration and dialogue. Keep to one short line per clip. All characters here are fictional.
Prompt 1: late-night confession (this one has a real clip, shown below)
Modern bedroom at night, soft warm lamplight. A woman and a man in their mid-30s lie in bed facing each other, close but not touching. She looks at him with fear in her eyes, takes a breath and says quietly, "I'm scared you're going to leave me." He holds her gaze, then gently takes her hand. Slow push-in from a wide two-shot to a close-up of her face. Intimate, natural performances, realistic skin, soft shadows, quiet room tone, no subtitles.
This is the exact prompt behind the clip below, made with Seedance 2.0, 16:9, 8 seconds. It practices one emotional beat and one quoted line. Use Seedance 2.0 for faces and lighting.
Prompt 2: rainy bus stop
A fictional young man in a green rain jacket waits at an empty bus stop at night, rain streaking the glass shelter. An older woman with a yellow umbrella sits beside him and says warmly, "The last bus left an hour ago." He laughs, surprised, and says, "Then I'm glad you stayed." Static medium two-shot, cool street light, soft rain and distant traffic. No subtitles, no on-screen text.
It practices two short lines with a clear speaker order. Try Seedance 2.0, 8 seconds, 16:9.
Prompt 3: kitchen goodbye
A woman in a gray cardigan stands in a small kitchen holding a packed suitcase, morning sun on the counter. Her brother, in a white T-shirt, leans on the doorframe and says quietly, "Call me when you land." She smiles and nods without speaking. Slow push-in from a wide shot to a close-up of her eyes. Warm window light, kettle hum, no music, no text.
It practices a silent reaction after a line. Use Seedance 2.0 at 8 seconds, 16:9.
Which prompts suit product ads?
Seedance 2.0 prompts for ads need a clean surface, one product and a slow camera. Use Mini to test the framing, then spend credits on the final. Describe the object, not a brand.
Prompt 4: perfume bottle on stone
A tall frosted-glass perfume bottle with a gold cap stands on a wet black stone slab. Water beads roll down the glass. Slow 180-degree orbit, macro lens, shallow depth of field. Soft white studio light with a thin rim light, tiny mist particles drifting. Quiet glass-clink sound. Clean frame, no logos, no text.
It practices one object and one orbit. Start on Mini at 4 seconds, 1:1, then finish on Seedance 2.0 at 8 seconds.
Prompt 5: pocket speaker on a desk
A small cylindrical speaker in matte olive fabric sits on a sunlit wooden desk beside a plant and a notebook. A hand taps the top and the speaker lights a soft white ring. Slow push-in from medium shot to close-up. Warm morning light, gentle lo-fi beat that pulses with the light. No logos, no text, no subtitles.
It practices linking sound to an on-screen action. Use Mini for drafts, 8 seconds, 16:9.
Which Seedance 2.0 prompts work for vertical social video?
Vertical Seedance 2.0 prompts need a centered subject, because feeds crop the edges. Set 9:16 in the generator as well as in the text. Mini is a good fit, since these are often quick, repeatable posts.
Prompt 6: handmade mug (real clip below)
Cinematic, ultra-realistic vertical video in a warm, rustic pottery workshop. Wet clay spins on a pottery wheel as skilled hands shape it into a tall ceramic mug and smooth the rim with a sponge. Close-up of fingers pressing a handle into place, then hot coffee is poured into the finished mug with gentle steam. Warm natural window light, realistic clay texture, detailed hand movement, shallow depth of field, soft wheel hum, no text, no logos.
This exact prompt was rendered with Seedance 2.0 Mini, 9:16, 8 seconds. It practices a process with a clear start and finish.
Prompt 7: outfit change transition
Vertical 9:16 video. A fictional woman in a plain white tee stands centered in a bright apartment hallway. She snaps her fingers and her outfit changes to a long red coat and black boots, the hallway light turning warmer. Camera locked at chest height, subject centered. Upbeat pop beat with a sharp snap sound. No text, no logos.
It practices one clear transformation beat. Use Mini, 4 to 8 seconds, 9:16.
Which prompts fit dance and music clips?
Seedance 2.0 prompts for dance work when you give beats and a camera that stays out of the way. Name the rhythm: "on three sharp beats". Ask for no text so the frame stays clean.
Prompt 8: rooftop dance teaser (real clip below)
Cinematic debut teaser for an original virtual K-pop girl group. Five performers in chrome-accented black outfits stand in a staggered diagonal on a futuristic rooftop arena at night, violet laser beams slicing through haze. On three sharp beats they snap into synchronized poses, then break into fast, sharp choreography with body waves and a 180-degree turn. The camera glides sideways, then orbits half a circle. LED panels pulse blue and violet. Energetic electronic beat, no text.
This is the exact prompt behind a clip made with Seedance 2.0, 16:9, 8 seconds. It practices synchronized motion and light timed to music.
Prompt 9: subway busker
A fictional street musician in a denim jacket plays an acoustic guitar on a subway platform, a few commuters swaying nearby. Camera glides in a slow arc from left to right, then pushes to a close-up of his hands on the strings. Warm tiled-wall light, train rumble in the distance, bright fingerpicked melody. No singing, no text.
It practices a single instrument with room sound. Use Seedance 2.0 at 8 seconds, 16:9.
Which Seedance prompts work for anime?
For anime, put the style word first, because it carries the whole look. "Hand-drawn anime" plus a color palette is usually enough.
Prompt 10: train at dusk
Hand-drawn anime style, a schoolgirl with a red scarf sits by the window of an empty commuter train at dusk, fields sliding past in orange and pink. Slow push-in on her face as she smiles and whispers, "Almost home." Soft light flares, gentle piano, rhythmic rail clicks. No subtitles, no on-screen text.
It practices a stylized close-up with one line. Use Mini for drafts, 8 seconds, 16:9.
Which prompts suit travel and landscapes?
Landscape Seedance 2.0 prompts need scale, time of day and moving air or water. Choose one big camera move and let nature supply the rest.
Prompt 11: terraced rice fields
Aerial shot rising slowly over green terraced rice fields at sunrise, mist filling the valleys between the ridges. A farmer in a straw hat walks a narrow path below, tiny in the frame. Soft golden backlight, long shadows, birdsong and a light breeze. Calm, documentary look, no text.
It practices scale with a small human figure. Use Seedance 2.0, 12 seconds, 16:9 or 21:9.
Prompt 12: harbor town walk
A fictional traveler in a beige linen shirt walks along a cobbled seaside street in a small harbor town at golden hour, fishing boats bobbing behind him. Tracking shot at walking pace from the side. Warm sunlight on whitewashed walls, gulls, rope creaks and distant church bells. Relaxed travel-film mood, no text.
It practices a smooth follow shot. Use Mini, 8 seconds, 16:9.
How do you prompt food shots?
Food Seedance 2.0 prompts need texture words and sound: sizzle, steam, pour. Keep the hands and the dish in focus.
Prompt 13: ramen pour
Close-up of a cook in a small ramen shop ladling steaming broth over fresh noodles in a black bowl, then placing a soft egg and green onions on top. Slow push-in, shallow depth of field. Warm tungsten light, rising steam, broth splash and quiet kitchen chatter. Realistic food texture, no text, no logos.
It practices steam, liquid and slow detail. Use Mini, 8 seconds, 1:1 or 9:16.
Prompt 14: sourdough slice
A baker slices a golden sourdough loaf on a flour-dusted wooden board, the crust crackling as the knife cuts through. Steam rises from the open crumb, and butter melts on a warm slice. Macro shot, then a slow pull-back to the bakery counter. Soft morning window light, crisp crust crackle, no music, no text.
It practices texture and a macro-to-wide move. Use Mini, 4 to 8 seconds, 4:3.
How do you write a multi-shot Seedance 2.0 prompt?
For multi-shot Seedance 2.0 prompts, number the shots and keep subject, light and style the same across them. The Seed team's launch post highlights 15-second multi-shot output, so this is the format to learn. Aim for two to four shots, and give each a different job.
Prompt 15: harbor mystery (12 seconds)
Shot 1: wide shot of a foggy fishing harbor at dawn, a small blue boat rocking at the pier, a lone woman in a yellow raincoat walking toward it. Shot 2: medium shot from behind her as she steps onto the boat and finds a sealed wooden box on the deck. Shot 3: close-up of her hands lifting the lid, a soft golden glow lighting her face. Same woman, same raincoat, same gray-blue light throughout. Foghorn, water lapping, a low rising hum. No text.
It practices three clear beats. Use Seedance 2.0, 12 seconds, 16:9.
Prompt 16: product reveal in four shots (15 seconds)
Shot 1: close-up of a closed black notebook on a white desk, a hand resting on it. Shot 2: the hand opens the cover, revealing a hand-drawn rocket sketch. Shot 3: the sketch lifts off the page as a tiny paper rocket and flies across the desk. Shot 4: slow pull-back as the rocket lands on a window sill in sunlight. Same desk, same warm light, soft paper-rustle sounds and a light whoosh. Playful mood, no text, no logos.
It practices a transition from real to imagined. Use Seedance 2.0, 15 seconds, 16:9.
How do you write Seedance 2.0 prompts for image to video?
For image to video, Seedance 2.0 prompts should describe what changes, not what the picture already shows. The image sets the first frame, and an optional second image sets the last frame. So the prompt only needs motion, camera, light shifts and sound.
- Weak: "A woman in a red coat on a snowy street."
- Better: "She turns toward the camera and smiles. Snow begins to fall. Slow push-in. Soft wind, distant bells."
Use illustrations, product shots or AI-generated characters as your first frame. Seedance 2.0 does not accept reference images with real faces. Our Seedance image to video guide covers first and last frames in detail.
What common mistakes weaken a prompt?
Most weak Seedance 2.0 prompts fail in the same few ways, and weak Seedance 2.0 prompts are easy to fix. Fix one at a time.
| Mistake | Why it hurts | Fix |
|---|---|---|
| Three camera moves in one shot ("pan, zoom and orbit") | The moves fight each other | Pick one move per shot, or split into shots |
| A pile of adjectives ("stunning, epic, beautiful, amazing") | Adds no visible detail | Replace with one concrete detail, such as "low golden light" |
| No sound | Audio is left to chance | Add ambience, music or a quoted line |
| Excluded things written as scenes ("a street with no cars") | Names the thing you want gone | Say "empty street" or add "no text, no logos" |
| Several actions at once | Motion gets muddy | One verb per shot |
| Real people or famous characters | Real faces are restricted, and copyright issues apply | Use fictional people and original designs |
| Very long prompts | Harder to debug | Stay near 50 to 120 words |
How do you use these Seedance 2.0 prompts in our studio?
Copy a prompt, paste it into the Seedance video generator, and test it on Mini at 4 seconds and 480p first. That test is 24 credits, and generating needs a subscription. See pricing and our Seedance 2.0 Mini guide for the model differences.
These Seedance 2.0 prompts run on both models in our studio through the WaveSpeed API. The prompt box takes up to 3,500 characters, which is a limit on our site, not the model. Audio is on by default. Aspect ratios are 16:9, 9:16, 1:1, 4:3, 3:4 and 21:9.
If the first pass is close, edit your Seedance 2.0 prompts by changing one slot of the formula and try again. When it works, switch to Seedance 2.0 for the final clip. For costs, see our Seedance 2.0 pricing guide.
FAQ
Are these Seedance 2.0 prompts free to use?
Yes, you can copy and edit every prompt here for free. Generating clips in our studio needs a subscription and costs credits per second.
How long should a Seedance prompt be?
BytePlus recommends no more than 1,000 English words or 500 Chinese characters. Most good Seedance 2.0 prompts are far shorter, around 50 to 120 words.
Can I use celebrities or cartoon characters?
We do not recommend it. The Seedance 2.0 series does not accept reference images or videos with real faces, and protected characters raise copyright problems. Invent your own characters.
Do these Seedance 2.0 prompts include sound?
Yes. Seedance 2.0 generates audio with the video, so each prompt names ambience, music or a quoted line.
Do all 16 Seedance 2.0 prompts have finished clips?
No. Three have clips made in our studio, shown above. The other thirteen are starting points we have not published results for.
Can I try them on the free plan?
Yes, within the 24 starter credits. See our free Seedance 2.0 guide for ways to stretch them.
Sources checked Nov 14, 2026: BytePlus Seedance 2.0 docs, Seed launch post, Dreamina Seedance 2.0 page.