Seedance 2.5 API — Cinematic AI Video, Up to 30 Seconds
Generate multi-shot videos with native audio from a text prompt, an image, or a reference clip. One Seedance 2.5 API call, one render, ready to publish.
From $0.07 per second — 480p with a reference clip
What is the Seedance 2.5 API?
The Seedance 2.5 API is a developer-friendly endpoint for generating cinematic AI video with ByteDance's Seedance 2.5 model. Send a text prompt, a still image, or a reference clip, and get back a finished video — multiple shots, synchronized native audio, and clips up to 30 seconds — from a single request.
It runs the same model that powers the live generator on this page, so you can storyboard a shot in the browser and ship the identical workflow in code. Generation is asynchronous and simple: submit a job, poll until it is ready, then download a publish-ready MP4. There is no render farm to manage, no separate audio pass, and no stitching short clips together by hand.
Seedance 2.5 is built for people who need real output, not just demos — marketing teams, social creators, product studios, and app developers who want cinematic video on tap. Each render is billed per generation in credits with transparent per-second pricing, and new accounts get free starting credits so you can test with real prompts before wiring it into production.
One Seedance 2.5 API call can accept text, images, video, and audio references together, supports clips of 4 to 30 seconds, and returns multi-shot scenes with camera movement and sound already in place — no timeline editor required.
What Seedance 2.5 costs
Billed per second of finished video. No subscription, no minimum spend.
| Resolution | Generation only | With a reference video |
|---|---|---|
| 480p | $0.118589 | $0.071153 |
| 720p | $0.200118 | $0.120071 |
| 1080p | $0.346392 | $0.206928 |
A reference video bills its own length on top of the output, and any request containing video is charged a floor of five thirds of the output duration. Reference images and reference audio add nothing. Failed generations are refunded in full.
Rates read on 2026-09-24.
Everything the Seedance 2.5 API gives you
Cinematic video generation, built into one endpoint.
Thirty Seconds in One Take
A single generation runs up to 30 seconds with several shots inside it. Seedance 2.0 and most rivals stop at 15.
Native Audio, Built In
Dialogue, sound effects, and ambient tone generate with the video, so there's no syncing or stock-audio step.
Multi-Shot Storytelling
Describe a scene with several shots and Seedance 2.5 cuts between camera angles in a single render.
Text, Image & Video Inputs
Start from a prompt, animate a still image, or extend a reference clip without switching tools.
One API Call
Send a request to the Seedance 2.5 API, poll for status, and pull back a finished video URL.
Clips up to ~15 Seconds
Generate continuous scenes long enough to tell a beat, not just a three-second loop.
How it works
From input to finished clip in three steps.
Send your input
Pass a text prompt, image, or reference video plus your settings to the Seedance 2.5 API.
Seedance 2.5 renders
The model generates a multi-shot clip with synced native audio at your chosen resolution.
Download & ship
Get back a finished video URL, ready to post or drop into your app.
Built for real production work
Social & short-form
Spin up vertical clips with sound for TikTok, Reels, and Shorts in one render.
Ads & product demos
Turn a product photo into a moving spot with narration and music built in.
Storyboards & previs
Block out multi-shot scenes in 1080p before you commit a real crew or budget.
App & platform features
Embed the Seedance 2.5 API so your users generate video without leaving your product.
Seedance 2.5 vs Seedance 2.0
Seedance 2.5 builds on Seedance 2.0 with three real upgrades: one generation now runs up to 30 seconds where 2.0 stopped at 15, it accepts up to 30 reference images alongside reference clips and audio, and it can edit or extend footage you already have. Everything creators relied on in 2.0 — native audio, multi-shot scenes, and realistic motion — carries straight over. Both top out at 1080p.
If you already integrated Seedance 2.0, the 2.5 API keeps the same submit-and-poll workflow, so upgrading is mostly a model-id change. You get sharper, longer-lasting detail for premium ads and large-format screens without rewriting your pipeline.
Every Seedance 2.5 reference page
Parameters, rates and limits, written out per topic. Each page answers one question rather than summarising all of them.
Content limits and NSFW
Seedance 2.5 itself
Compared with alternatives
Frequently asked questions
The Seedance 2.5 API is an endpoint for generating cinematic AI videos from the Seedance 2.5 model. It accepts text-to-video, image-to-video, and reference-video requests, and returns multi-shot clips with native audio at up to 1080p. You submit a generation job, poll for its status, and download the finished video URL.
Yes. Native audio — dialogue, sound effects, and ambient tone — is generated together with the video in a single pass, so there is no separate syncing or stock-audio step. Put spoken lines in quotes in your prompt to guide the voice track, and the model matches sound to the on-screen action.
Seedance 2.5 renders up to 1080p and clips of 4 to 30 seconds, with multiple shots possible inside one generation. Common aspect ratios are supported — 16:9, 9:16, 1:1, 4:3, 3:4 and 21:9 — so you can publish to YouTube, TikTok, Reels, or wide cinematic formats without resizing.
You can start from a text prompt, animate a still image, set a first and last frame, or extend a reference video — and combine image, video, and audio references in one request to lock in brand, character, or style continuity. Text-only generation works too when you just want a scene from a description.
Send a POST request with your prompt and settings to create a generation task, poll the task endpoint until its status is completed, then download the returned video URL. It is the same asynchronous submit-poll-download pattern shown in the generator above; full request and response formats live in the API docs.
Yes. Use the live generator at the top of this page to create a video with free starting credits, no setup required. When you are ready to automate, the same model is available through the Seedance 2.5 API so your app can generate video for your own users directly.
Generation is billed per second of output, so a short 480p clip costs far less than a long 1080p render. Rates are published per resolution and shown on this page, and failed generations are refunded in full.
Seedance 2.5 extends a single generation to 30 seconds where 2.0 stopped at 15, accepts up to 30 reference images, and can edit or extend existing footage, while keeping native audio, multi-shot storytelling, and realistic motion. Both render up to 1080p. The API workflow is unchanged, so moving from 2.0 to 2.5 is mainly selecting the new model.
Latest from the blog
Guides, comparisons and real prices for Seedance 2.5
Cheapest Platform for Seedance 2.5: How to Compare
How Much Does Seedance 2.5 Cost? Full Rate Breakdown
How to Access Seedance 2.0: From Nothing to First Clip
How to Access Seedance 2.5: Key, Request, First Render
How to Use Seedance 2.0 for Free: An Honest Accounting
