Picture and sound in one pass
Dialogue, breathing, footsteps and room tone are generated with the video, in sync. On Affogato, sound is on by default and turning it off does not lower the price.
Seedance 2.0 · by ByteDance
Seedance 2.0 AI video generator
Seedance 2.0 is ByteDance's video model, known for people who act and speak in the same pass as the picture, motion that follows physics, and taking several reference images at once. Write a scene below and it opens in Video Studio on Seedance 2.0.
Example video
Opens in Video Studio with the model you picked. You see the price in credits before anything runs.
What Seedance 2.0 is known for
Motion that follows physics
ByteDance built 2.0 around motion that obeys physical laws: things fall, splash and shatter with weight. Reviewers single it out for action and motion realism.
Reference-first
Give it up to nine reference images (up to four characters, three images each) or a reference clip, and refer to them in the prompt. Product shots and recurring characters stay recognisable.
Any length from 4 to 15 seconds
You pay for the seconds you ask for, not a fixed bucket, and one generation can cut between several shots.
Where it's weaker
- Photos of real people are often refused as references, a rule ByteDance added in March 2026.
- Long lines of dialogue drift out of lip sync; keep each spoken line short.
- Text and logos inside the frame are unreliable.
- If the prompt says nothing about sound, it may add music. Write "no music" when you mean it.
- Renders at 720p by default; 1080p costs about twice as much.
- Maker
- ByteDance Seed · launched February 2026
- Clip length on Affogato
- Any length from 4 to 15 s
- Resolution
- 720p · 1080p (Fast: 720p)
- Aspect ratios
- 9:16 · 3:4 · 1:1 · 4:3 · 16:9 · 21:9
- Sound
- Native audio, on by default, same price either way
- Inputs
- Text · start and end frame · up to 9 reference images · reference clips of 2–15 s
Made with Seedance 2.0 on Affogato
Every clip below was generated with the prompt shown, unedited. Load any prompt into the generator and make your own version.
Acting
A letter by the window
Prompt
Close-up, vertical frame: a woman in her sixties sits by a rain-streaked window reading a handwritten letter. Her eyes move along the lines, her lips press together, then her chin trembles as tears well up. She lowers the letter, exhales a shaky laugh and whispers, "you remembered." She smiles through the tears. Soft grey window light, shallow depth of field, the camera holds almost still with a very slow push-in. Sound: rain on glass, paper rustling, her quiet breath.
Dialogue and lip sync
Street interview, one line
Prompt
Handheld vertical street interview, phone-camera style. A young man with curly hair and a denim jacket stands on a busy city sidewalk, holding a coffee. Someone off-camera asks what he loves about mornings. He laughs, thinks for a second and says, "honestly? the ten minutes before anyone texts me." He shrugs and sips his coffee. Natural daylight, passers-by blurred behind him. Sound: city traffic, his natural voice, a faint laugh from the interviewer.
Natural motion
Dancer in a sunlit studio
Prompt
Full-body wide shot, the dancer always in frame from head to toe: a contemporary dancer in a loose white shirt and wide linen trousers performs alone in a sunlit wooden studio. She turns into a slow spin, arms sweeping out, then drops low into a controlled floor roll and rises in one fluid motion. The fabric of her shirt trails each movement. The camera circles her slowly at a distance. Dust floats in shafts of afternoon light. Sound: soft piano, bare feet sliding on the wooden floor.
Physics
Wine glasses hit the floor
Prompt
Slow motion, vertical frame: a waiter in a busy restaurant trips and a tray of six wine glasses tips forward. The glasses tumble, red wine arcs through the air in ribbons, and the glasses shatter on the tiled floor, shards scattering and wine splashing outward. Diners in the background turn their heads. Warm evening light, shallow depth of field. Sound: a gasp, glass shattering, a stunned hush.
Multi-shot
Rooftop chase in three shots
Prompt
A night-time rooftop chase in three shots. Shot one, wide: a courier in a red jacket sprints across a wet rooftop, a security guard with a flashlight running after her. Shot two, low tracking shot: she vaults a ventilation duct and lands in a roll, splashing through a puddle. Shot three, close-up: she stops at the roof edge, breathing hard, and looks back over her shoulder with a grin. City lights below, light rain, neon reflections. Sound: footsteps on wet gravel, heavy breathing, distant sirens.
Image reference
A pitcher from one reference photo
Prompt
A vertical product ad featuring the white ceramic pour-over kettle from @Image1. The kettle sits on a light oak counter in a bright minimalist kitchen. Hot water pours from its spout in a thin steady stream into a glass coffee dripper, steam curling up. The camera slowly pushes in and tilts down to finish on the kettle's curved handle. Keep the kettle's exact shape, color and matte finish. Morning sunlight, soft shadows. Sound: pouring water, a quiet morning room tone.
Fur and water
Retriever shaking off the lake
Prompt
Slow motion, vertical frame: a wet golden retriever stands on a sunny lakeside dock and shakes off water. Its fur ripples in waves from head to tail, ears flapping, and droplets spray out in a glittering arc against the backlight. It finishes, looks straight at the camera and wags its tail. Golden hour, lake shimmering behind. Sound: water spray, a happy pant.
Fire and food
Noodles in a flaming wok
Prompt
Close-up in a busy restaurant kitchen: a chef tosses noodles and vegetables in a blackened wok over roaring flames. The noodles lift, flip and fall back into the pan, sauce glistening, as a burst of fire flares up the side of the wok. Her other hand shakes in a pinch of sesame seeds. Warm firelight on stainless steel, steam everywhere. Sound: sizzling oil, the clang of the wok, a whoosh of flame.
Which version to pick
Start here
Seedance 2.0
Most clips
Text, a start frame or references in, a clip with sound out. Start here for anything up to 15 seconds.
5 s clip 275 credits · 4–15 s
Open with Seedance 2.0Option 2
Seedance 2.0 Fast
Drafts and quick takes
The same model tuned for speed, at 720p and a lower price. Find the shot here, then re-run it on 2.0.
5 s clip 220 credits · 4–15 s
Open with Seedance 2.0 FastOption 3
Seedance 2.5
Clips longer than 15 seconds
The newer version: up to 30 seconds and a cheap Draft mode you can enhance later. It costs more per second.
5 s clip 429 credits · 4–30 s
Open with Seedance 2.5| Seedance 2.0 | Seedance 2.5 | |
|---|---|---|
| Released | February 2026 | June 2026 (wider launch in July) |
| Clip length | 4–15 s | 4–30 s |
| Resolution on Affogato | 720p · 1080p | 480p · 720p · 1080p |
| Fast tier | Yes | No |
| Draft, then enhance | No | Yes |
| References | Up to 9 images, or clips | More at once (ByteDance says up to 50) |
| Price for the same length | Lower | Higher |
ByteDance says 2.5 follows prompts about 20% better. For a clip of 15 seconds or less, 2.0 costs less; choose 2.5 when the shot has to run longer or you want to draft cheaply first.
Seedance 1.5 Pro and 1.0 Pro are still in the studio's model menu: much cheaper, up to 12 seconds, and 1.0 has no sound. Credits are read from the live catalog; the Standard plan ($24/month) includes 2,000 credits, about 7 five-second Seedance 2.0 clips.
Prompting Seedance 2.0 well
- Write a shot brief, not keywords: who is in the frame, what they do, then the place, the look and the camera. Skip "8k" and "masterpiece".
- Use verbs and named camera moves (push-in, dolly, locked-off). Give physical action a consequence: glass shatters, water sprays, dust kicks up.
- Put dialogue in double quotes, keep each line short, and say how it is spoken ("whispers", "laughs and says").
- Name the sounds you want and write "no music" when you don't want a score.
- For several shots, write "Shot 1:", "Shot 2:" or "cut to". About three cuts per clip works best.
- With references, tag them in the prompt (@Image1, @Video1) and say what each one is for: the product, the character, the camera move.
- Test at 5 seconds and 720p, then render the winner longer or at 1080p.
From prompt to clip
Describe the scene or add a photo
Write the shot in the box at the top of this page, or start from an example. Add a photo to use it as the first frame.
Sign in and land in Video Studio
Your prompt, photo, format and the Seedance version you picked carry over. Nothing runs until you press Generate.
Pick the length, then generate
Choose any length from 4 to 15 seconds. The credit price shows under the Generate button before anything is spent.
Add references when it has to match
Switch to Reference to Video to bring in a product, a character or a camera move from your own images or clips.
Questions about Seedance 2.0
What is Seedance 2.0 best at?
People who move and speak with their sound generated in the same pass, motion with believable weight, and keeping a product or character recognisable from reference images. It is weaker on text in the frame and long lines of dialogue.
What is the difference between Seedance 2.0 and 2.5?
Seedance 2.5 runs up to 30 seconds instead of 15, takes more references at once and adds a Draft mode, but costs more per second and has no Fast tier. For clips of 15 seconds or less, 2.0 is the cheaper choice.
Is Seedance 2.0 better than Kling 3.0?
They are good at different things. In our same-prompt tests Seedance 2.0 won on acting and on a three-shot sequence, Kling 3.0 on following a detailed shot list, and Kling costs less for the same length. The comparison page shows every test side by side.
Does Seedance 2.0 generate sound?
Yes. Dialogue, effects and ambience come with the picture. On Affogato sound is on by default and the price is the same with it off.
Can I use a photo of myself or a real person?
Often not. Since March 2026 ByteDance blocks most reference images that show real faces, so those runs are refused. Illustrated or generated characters, products and places work.
Is Seedance 2.0 free on Affogato?
No. Every run costs credits, priced by the second and shown before you generate. Credits come with a monthly plan and unused credits roll over up to the plan's cap.
Can I use Seedance 2.0 videos commercially?
Yes. Paid Affogato plans include a commercial license and no watermark, within our terms of service and ByteDance's content rules.
Can I call Seedance 2.0 from code?
Yes. The Seedance versions on Affogato are also available through the Affogato API and MCP server, at the same credit price.
Keep exploring
Facts checked November 14, 2026. Sources: ByteDance Seed: Seedance 2.0 launch · fal: Seedance 2.0 API · fal: Seedance 2.0 prompting guide · fal: Seedance 2.5 vs 2.0 · The Next Web: Seedance relaunch rules · The Next Web: Seedance 2.5 · Akita on Rails: Seedance 2.0 first impressions