sedance.app

Seedance 2.5 · API live

Seedance 2.5:
thirty seconds in one pass.
No seam to hide.

The Seedance 2.5 AI video generator makes 30 continuous seconds in a single pass — scene changes, tempo shifts and camera moves inside one unbroken take, not stitched and not extended. Generate it in the browser, or with one API call.

30s
single pass
50
references
4K
unverified

See what a single pass looks like

One continuous take with synchronized audio — no cut, no stitch, no seam to hide. Press play with the sound on: the audio is generated with the video, not laid over it afterwards.

Share this

Preview reel: a pilot in a cockpit lit by green heads-up displays, the camera holding on their face as the interface flickers around them — one continuous cinematic take with synchronized ambient audio.

Price it before you spend

Plans from $18/mo
Live generator · exact cost, before you commitno watermark on any plan
References

0/50 references · up to 40 images, 5 videos (≤15s each), 5 audio

Name them in the prompt the way the model does — “match the camera move in @video1”.

Model

30 seconds in a single pass — no stitching, no seam to hide.

Resolution
Aspect
Tier
Duration — 4s
4s30s
00:00:04:00480p16:9standard267cr≈ $2.67
Balance0cr
After this generation-267cr
Typical wait210s

What is Seedance 2.5?

Seedance 2.5 is ByteDance’s video-generation model, live on the BytePlus ModelArk API since 16 July 2026. It generates up to 30 seconds of continuous video in a single pass — scene changes and tempo shifts inside one take — and accepts up to 50 multimodal references. This site resells it with the per-second cost shown before you generate.

Earlier Seedance models — 2.0, 2.0 Fast, 2.0 Mini and 1.5 Pro — top out at 15 seconds and reach longer runs by stitching or extending, which is where continuity usually breaks. 2.5 removes the join entirely. That is the whole reason to pay more per second for it.

Model family
Seedance (ByteDance)
Longest single pass
30 seconds
Resolutions
480p · 720p · 1080p · 4k
Multimodal references
50
Native audio
Yes
Typical wait
210s (p50)
30s at 1080p
7,179 credits · $71.79

Figures ByteDance has announced but we have not yet confirmed against the API are marked unverified on the model page rather than repeated as fact here.

What Seedance 2.5 adds — and what we could actually verify

01verified

30 seconds, one pass

4–30 seconds generated as a single continuous take, with scene changes and tempo shifts inside it. Not stitched, not extended — there is no seam to hide.

02unverified

Up to 50 references

50 multimodal inputs in one generation: 40 images, 5 video clips and 5 audio tracks, against roughly a dozen on 2.0.

03unverified

Region-level editing

Change one element and leave the rest of the frame untouched, instead of regenerating the whole take and hoping the rest survives.

04unverified

R2V motion control

Green-screen and white-model references drive motion, spatial layout and character interaction, so generation follows a structured visual plan rather than text alone.

05unverified

3D camera blockout

Plan the camera move against a white-box preview before spending a generation on it. The cheapest take is the one you did not need to run.

06unverified

Multilingual dialogue

Dialogue in multiple languages with lip movement synced to it, carried by the model rather than added in post.

Every figure marked unverified is ByteDance’s own announced claim, which we have not yet confirmed against the provider API. No independent benchmarks for Seedance 2.5 existed when this page was written — so rather than restate a keynote, we mark the gap and publish measurements as they land.

Every clip here is ours

Every take on this page was generated with Seedance here — no stock, no cherry-picked vendor reel. Prompt, model and settings shown on each one.

All 4Seedance 2.5 430s takes 4With references 4
Still from an AI-generated video: Smooth gimbal walkthrough at chest height, moving forward through an open-plan living room in late-afternoon light10s · 1080p

Smooth gimbal walkthrough at chest height, moving forward through an open-plan living room in late-afternoon light. The camera passes a linen sofa on the right, tracks past a dining table, and settles looking out through floor-to-ceiling windows onto a garden. Warm low sun, long shadows across pale oak flooring, no people. 24mm lens, no distortion. Quiet room tone, faint birdsong outside.

Seedance 2.01777cr
Still from an AI-generated video: Aerial shot beginning close on a lone cyclist on a coastal road at sunrise, then rising and pulling back to reveal a long headland, sea mist in the bays and the sun breaking over the ridge behind10s · 1080p

Aerial shot beginning close on a lone cyclist on a coastal road at sunrise, then rising and pulling back to reveal a long headland, sea mist in the bays and the sun breaking over the ridge behind. Continuous smooth ascent, no cuts. Golden low sun, long shadows, natural colour. Wind and distant surf, no music.

Seedance 2.01777cr
Still from an AI-generated video: Extreme macro, slow motion6s · 1080p

Extreme macro, slow motion. Dark chocolate sauce pours in a thin ribbon onto a scoop of vanilla ice cream in a ceramic bowl, catching and pooling in the crevices. Hard backlight through the pour so the sauce glows at its edges, dark background falling to black. 100mm macro lens, very shallow focus on the point of contact. Close, syrupy pouring sound.

Seedance 2.01069cr
Still from an AI-generated video: Studio product shot on a seamless charcoal backdrop8s · 1080p

Studio product shot on a seamless charcoal backdrop. The camera orbits slowly clockwise around a matte-black wireless speaker at desk height, 35mm lens, shallow depth of field. A large softbox key light from the upper left rolls a soft specular highlight across the cylinder as it turns; a cool rim light separates it from the background. No text, no hands. Subtle low-frequency room tone.

Seedance 2.01423cr

Our numbers, not the keynote's

Full benchmark

Every Seedance 2.5 capability figure in circulation is ByteDance’s own. Our benchmark measures what actually comes back — billed cost, latency, failure rate, and how often a 30-second take holds character end to end. It has not been run yet. The rates below are the published vendor figures our pricing currently rests on, not measurements, and the columns we cannot honestly fill are marked as such.

RouteBilled /s 720p30s clip 720pp95 latencyFailure
BytePlus ModelArkunverified$0.3375$10.13not measurednot measured
fal$0.4296$12.31not measurednot measured

Billing shapes differ — some routes charge for input-video seconds, some add a per-request token component — so headline per-second rates are not comparable until they are normalised. These figures are.

For developers

One API call. One credit balance. Two front doors.

Idempotency keys on create, rate-limit headers on every response, a stable error code you can switch on, and a status page publishing real per-model p50 and p95. Prepaid packs never expire.

$0.012/credit6 models4 languagesDocs →
curl https://sedance.app/api/v1/generations \
  -H "Authorization: Bearer sk_live_..." \
  -H "Idempotency-Key: $(uuidgen)" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedance-2.5",
    "duration_sec": 30,
    "resolution": "1080p",
    "prompt": "one continuous take, camera rising over a night market",
    "confirm_cost": true
  }'

→ 202 { id, status: "reserved",
        credits_held: 7179, eta_seconds: 210 }

Seedance questions, answered

What is Seedance 2.5?
Seedance 2.5 is ByteDance’s video-generation model, available on the BytePlus ModelArk API since 16 July 2026. Its headline capability is 30 seconds of continuous generation in a single pass — including scene changes and tempo shifts inside one take — where earlier Seedance models stitched or chained extensions.
How long can a Seedance 2.5 video be?
30 seconds in one pass, generated 4–30 seconds per request. That is a single continuous take, not several clips joined together, which is what separates 2.5 from every earlier Seedance model.
How much does a 30-second Seedance video cost?
A 30-second clip costs roughly six times a 5-second one, because video models bill per output second. Our cost calculator shows the exact credit price for every duration and resolution before you generate, and the studio asks you to confirm any job over 800 credits.
Is there a free Seedance AI video generator?
Not here. Every generation costs us real GPU time — a 30-second clip is dollars, not cents — so there is no free tier and no trial credits. Plans start at $12/month, and credit packs start at $20 and never expire. You can price any generation exactly before you spend anything.
Seedance 2.5 vs Seedance 2.0 — which should I use?
Use 2.0 or 2.0 Fast to draft: they are cheaper per second and cap at 15 seconds. Move to 2.5 when you need a single unbroken take longer than that, more reference inputs, or the region-edit and reference-to-video controls. Drafting short first costs less per usable clip.
Is Sedance the same as Seedance?
Yes. Sedance, Seedance, SeedDance and Seed Dance are all spellings of the same ByteDance model family. You are in the right place.
Can I use Seedance videos commercially?
Paid plans include commercial use, subject to the upstream provider’s terms and your holding the rights to anything you upload as a reference. Free-tier output is watermarked and not licensed for commercial use.
Is there a Seedance API?
Yes. The same credit balance powers the web studio and the REST API — one balance, two front doors. Keys are issued from the dashboard, the API is versioned at /api/v1, and it honours Idempotency-Key on create.
267cr≈ $2.67