One workspace for every current Seedance model. Write a prompt, add images, clips or audio, choose Seedance 2.5 for 30-second reference-driven shots or Seedance 2.0 for 4K, and generate with native audio. Seedancee is an independent platform that runs the models through Kie.ai and keeps every result in your library.
Start a generation
What you can make
Seedance is a family of video generation models from ByteDance Seed. On Seedancee you can use it for product and UGC-style ads, fashion and food clips, real-estate walkthroughs, cinematic shorts, original animation, social hooks, music visuals, consistent characters, logo motion beds and travel footage. Output ranges from a 4-second 480p draft on Seedance 2.0 Mini to a 30-second 1080p one-take on Seedance 2.5 or a 15-second 4K hero shot on Seedance 2.0, with sound generated in the same pass.
Text to video: describe the shot and let the model build it from scratch.
Image to video: animate a first frame, optionally toward a last frame, keeping your source intact.
Reference to video: attach images, clips and audio for consistent identity, style, motion and rhythm.
Edit and extend (Seedance 2.5): revise a finished generation or continue it, up to twice.
Native audio: ambience, foley and simple music cues from the prompt, or your own reference audio.
How it works
01
Choose a model and mode
The model picker reads the live registry; only options the selected model supports are shown, so you cannot configure something that will fail.
02
Write the prompt and add media
Subject, action, camera, light, sound. Upload frames or references; limits are enforced before you spend anything.
03
See the price, then generate
The credit cost updates live. Generating reserves those credits; a failed job releases them automatically.
04
Review, remix, publish
Results land in your library with their prompt and settings. Remix into a new generation, extend it on 2.5, download, or publish to your own gallery page.
Choose the right Seedance model
All active models are listed below with capabilities taken from the model registry, which we verify against Kie.ai's documentation on the date shown in the last row. Seedance 2.5 is the default; the others are there because each does something 2.5 does not.
Active Seedance models on Seedancee
Seedance 2.5По умолчанию
Seedance 2.0
Seedance 2.0 Fast
Seedance 2.0 Mini
Seedance 1.5 Pro
Лучше всего для
Longer reference-driven scenes up to 30s with native audio and editing workflows.
High-quality 4–15s shots with up to 4K output and multimodal references.
Quicker iterations at 480p/720p with the same reference inputs as 2.0.
Lowest-cost drafts and social clips with strong visual consistency.
Previous-generation text and image-to-video with a fixed-lens option.
Fixed-lens option, lower cost than 2.0 at 1080p (1080p needs Pro)
Starter runs its models at up to 720p and 10-second clips, without reference videos. Pro includes every model at up to 1080p (4K on Seedance 2.0) and each model's full length, with reference videos, edit and extend. Options your plan does not include are shown with a lock rather than hidden.
Seedance 2.5
30-second reference-driven generation with native audio.
The fastest path from idea to clip. Works on every model. Best results come from one subject, one action and one camera move per 5–10 seconds, with the sound described in its own sentence.
Image to video
Upload a first frame and optionally a last frame. The generation starts from your pixels, which is why it is the workflow for products, interiors and illustrations that must stay recognisable. Seedance 2.0 and 2.5 can return the final frame for chaining shots.
Reference to video
Attach reference images for identity and style, clips for motion and audio for rhythm. Seedance 2.5 accepts up to 30 images, 10 videos and 10 audio files; the 2.0 variants accept 9, 3 and 3. This is how you keep a character consistent across a sequence.
Edit and extend
Available on Seedance 2.5 for finished generations: change an element with a new instruction, or continue the action. ByteDance states a 2.5 clip can be extended twice. Each edit or extension is a separate, quoted generation.
Order the prompt like a shot description: subject and setting, action, camera, light and look, sound.
Use concrete camera verbs: push-in, orbit, tracking shot, whip pan, locked off. Avoid stacking several moves in one short clip.
Put the sound last, in its own sentence, as layers: ambience, one or two foley events, optional music mood.
Draft at 480p on 2.0 Mini, then re-run the winning prompt on 2.0 (4K) or 2.5 (length, references).
For sequences, write beats in order and let 2.5 run them in one take, or chain last frame to first frame across generations.
Do not name real people, celebrities, copyrighted characters or third-party brands; it violates the content policy and produces unreliable results.
Example prompts
Cinematic one-take (Seedance 2.5)seedance-2.5
One continuous shot: a barista in a small corner cafe at opening time flips the sign to OPEN, walks behind the counter, pulls the first espresso as steam rises, then looks up as the first customer pushes the door open and the morning light floods in. Handheld, warm, 24fps. Sound: door bell, grinder, steam wand, quiet street outside.
A matte ceramic speaker on a walnut shelf; a slow orbit as a warm light sweep crosses the fabric grille, dust motes in the beam, dark background. Premium, minimal, no text. Sound: low room tone and a soft hum as the speaker's light ring turns on.
Vertical video. Whip pan from a blurry grey wall to a bright orange bicycle leaning against a graffiti mural in hard midday sun; the camera settles with a small bounce. Saturated, punchy. Sound: a whoosh, then street ambience.
Configuring the generator is free and needs no account. Generating needs a subscription, from $4.99 a week for Starter, and costs credits priced per model, duration, resolution and audio, with failed jobs refunded automatically. There are no free trial credits. Seedancee's REST API (/api/v1) uses the same pipeline, credits and refunds: it is part of the Pro plan and answers once API access is switched on, and until then it is in early access. The box below shows the current status.
Studio product film. A pair of matte-black over-ear headphones rests on a slab of wet dark slate. The camera begins on a tight macro of the ear cushion stitching, then pulls back on a slow motorized slider to reveal the full silhouette. A single hard key light from the upper left throws a crisp rim highlight along the headband; soft fill from a large overhead diffuser. Fine mist drifts through the beam. In the final second, a bead of water slides off the cup and lands with a soft, clean drip sound. Sound design: low room tone, subtle hum of the light, the single drip. No text, no logo.
Fashion film, 35mm anamorphic look, rich saturated color. A model in a floor-length emerald-green silk coat stands in a sunlit terracotta courtyard, late-afternoon golden light raking across ochre walls and a cobalt-blue door. A wind machine off-screen lifts the coat in slow waves. The camera orbits a quarter turn at hip height, shallow depth of field, soft warm lens flares. Palette: emerald, gold, terracotta, cobalt. Motion is unhurried; fabric moves in slow motion while the model holds a steady gaze into the lens. Sound: fabric rustle, distant wind, no music.
Cinematic scene. Interior of an overnight train at 2 a.m. A woman in a wool coat rests her forehead against the window; rain streaks the glass and the passing sodium lamps paint moving orange bars across her face. The camera is locked off on the seat opposite, slight handheld drift. Rich color: amber lamp glow against cool teal-blue night outside the window, deep shadows, subtle film grain, warm practical light from a reading lamp above. Halfway through she exhales and closes her eyes. Sound: steady rail rhythm, rain against glass, the lamp buzzing faintly.
Character close-up. An elderly lighthouse keeper with a white beard, weathered ruddy skin and a faded red wool sweater sits in a stone room, a kerosene lamp throwing warm amber light on the table while deep blue dusk fills the small window behind him. Medium close-up, 50mm, camera slowly pushing in. He looks just off lens, half-smiles, and says in a low gravelly voice: "They said the light would never fail. I never asked who they were." Lip movement matches the line. Lamp flicker, wind pressing on the shutters, a slow foghorn far away.
Vertical social ad, bright and punchy. A tall glass of ice sits on a sunlit white marble counter beside a bowl of bright oranges and a sprig of fresh mint, a coral-pink wall behind. Cold brew coffee pours in from above in slow motion, dark ribbons swirling through the ice, then a splash of oat milk clouds the glass from the bottom up. Quick whip-pan to a top-down shot as the milk blooms. High-key daylight, crisp highlights, very shallow focus. Energetic pacing with two fast cuts. Sound: ice cracking, liquid pour, one satisfying clink.
Abstract music visual, vivid color. Drops of magenta, cyan and gold ink fall into clear water inside a glass tank against a deep black backdrop, each color blooming and mixing into violet and teal clouds. Each drop lands on a slow 70 bpm pulse, tendrils unfolding in macro detail. The camera slowly rotates around the tank. Add a faint iridescent shimmer at the edges of the ink. High contrast, saturated, no text. Sound: a deep sub-bass pulse on each drop with soft reverb tails.
Configuring a generation and seeing its cost is free without an account. Running a generation needs a subscription and costs credits; there are no free trial credits. Plans start at $4.99 a week for Starter (350 credits per week, excluding tax).
Which plan do I need for each model?+
Starter includes Seedance 2.0 Mini, 2.0 Fast and 1.5 Pro at up to 720p and 10-second clips. Pro includes every model, adds Seedance 2.5 and 2.0, renders up to 1080p (4K on Seedance 2.0) and each model's full length, and adds reference videos, edit and extend.
Which Seedance models are available?+
Seedance 2.5 (default), Seedance 2.0, Seedance 2.0 Fast, Seedance 2.0 Mini and Seedance 1.5 Pro, all served through Kie.ai and verified on 29 September 2026. Seedance 1.0 is not offered because we have no verified endpoint for it.
How long can a Seedance video be?+
Up to 30 seconds on Seedance 2.5, 15 seconds on the 2.0 variants and 12 seconds on 1.5 Pro. Seedance 2.5 generations can be extended twice.
Does Seedance generate audio?+
Yes. Every active model generates native audio from the prompt when enabled, and Seedance 2.0 and 2.5 accept reference audio files.
Can I make 4K video with Seedance?+
Yes, with Seedance 2.0 on the Pro plan, for 4–15 second clips. Seedance 2.5 tops out at 1080p; the 2.0 Fast and Mini variants at 720p.
Is Seedancee the official Seedance website?+
No. Seedancee is an independent platform and is not affiliated with, endorsed by or sponsored by ByteDance. Seedance models are referenced to identify the third-party models we provide access to. ByteDance's official pages are linked in the sources.
Can I use the videos commercially?+
Yes, under Seedancee's terms and content policy, provided you hold the rights to any references you upload and your content does not infringe third-party rights.
Is there an API?+
Yes. Seedancee's REST API (/api/v1) creates generations with the same models, credits and automatic refunds as the workspace and sends signed webhooks. It is part of the Pro plan and answers once API access is switched on; until then it is in early access, which you can request through the contact page. The API page documents the endpoints and shows the current status. Kie.ai also offers hosted APIs for the same models.
