Seedance 2.0Text / First-Last Frame / Multimodal References

Seedance 2.0 Video Studio

Run Seedance 2.0 and Seedance 2.0 Fast with task creation, credit charging, and automatic result polling in one page.

2 Seedance Models

Standard and Fast variants share one submission flow.

3 Input Patterns

Prompt only, first-last frame, or multimodal references.

Auto Polling

Create taskId instantly and keep tracking until download.

Studio Preview

Seedance 2.0 Studio

Switch between clean text-only generation, first/last frame control, or multimodal image/video/audio references without building payloads manually.

Aspect Ratio

16:9 / 9:16 / 1:1

Resolution

480p / 720p

Duration

5 / 10 / 15s

References

Image / Video / Audio

Seedance 2.0 Video Generation

Supports text-to-video, first/last frame control, and multimodal reference generation with the existing task polling flow.

Create Task

First/last frame mode and multimodal reference mode are mutually exclusive, matching the model constraints.

Important Restrictions

Faces are not supported in input videos or images.
Input media must not exceed 720p.

Model

Aspect Ratio

Prompt

Resolution

Duration

First Frame

Last Frame

Reference Images

Optional, upload up to 9 image files or enter up to 9 public URLs.

Reference Videos

Optional, upload up to 3 video files or enter up to 3 public URLs. Total input video duration must not exceed 15s, and each uploaded file must be 50MB or less.

Reference Video Duration (s)

When video input is used, billing includes input video duration plus output duration. Total reference video duration must not exceed 15s.

Reference Audios

Optional, upload up to 3 audio files or enter up to 3 public URLs. Each uploaded file must be 10MB or less.

Public Visibility

Disable to keep the result private (VIP required).

Generate Audio

Enable soundtrack generation when the model supports it.

Web Search

Pass the web_search flag through to the underlying provider.

Return Last Frame

Request an extracted last frame result when supported.

Current Estimated Cost400

Current Rate80/s

Cost Breakdown

Output Duration5s

Input Video DurationNo video input

Billable Duration5s

Formula80 × 5 = 400

Task Results

Tasks are polled automatically. Preview and download after completion.

No tasks yet. Created tasks will appear here.

View History

Built For Seedance Workflows

One page connects prompt entry, asset upload, task creation, and result retrieval.

Mutual-Exclusion Guardrails
The form blocks invalid combinations before the request reaches the API.
Asset Upload Reuse
Local image, video, and audio files are uploaded to storage and converted into reusable asset URLs.
Shared History Pipeline
Generated tasks are saved into the same history and result polling system as other video models.

Why Use /seedance20

You get a production-ready Seedance entry instead of manually assembling low-level request payloads.

Fast Task Creation

Submit through the site API and receive taskId immediately.

Reference Flexibility

Switch between first/last frame control and multimodal image/video/audio references.

Automatic Refund Flow

Failed tasks stay inside the existing refund and synchronization path.

Private Delivery

VIP users can keep generations out of the public showcase.

Zero Payload Guesswork

The page maps UI inputs to the underlying request fields and lifecycle tracking automatically.

How To Use Seedance 2.0

Three steps from prompt to downloadable result.

01Step

Choose Model And Presets

Pick Seedance 2.0 or Fast, then set aspect ratio, resolution, and duration.

02Step

Add Frames Or References

Upload first/last frames or supply image, video, and audio references, but not both modes at once.

03Step

Poll And Download

Task status is polled automatically, then you can preview, copy URL, and download the final video.

Seedance Model Pricing

Per-second billing for Seedance 2.0 is based on resolution and whether video input is included.

Scenario	Credits	Capabilities
Seedance 2.0 · With video input	60/s	480p4-15s
Seedance 2.0 · Without video input	95/s	480p4-15s
Seedance 2.0 · With video input	125/s	720p4-15s
Seedance 2.0 · Without video input	205/s	720p4-15s
Seedance 2.0 Fast · With video input	45/s	480p4-15s
Seedance 2.0 Fast · Without video input	80/s	480p4-15s
Seedance 2.0 Fast · With video input	100/s	720p4-15s
Seedance 2.0 Fast · Without video input	165/s	720p4-15s

Feature Overview

Frontend form, backend task creation, and polling are wired together end to end.

Seedance-Specific API Route

Use /api/v1/video/seedance20 to create tasks and persist credit-backed history.

Unified Result Tracking

Reuse /api/video-generations/check-result for task progress and result URLs.

Multimodal Input Mapping

Map images, videos, and audios into the correct multimodal reference fields.

Manual Download Ready

Download locally stored result videos or copy direct links after completion.

FAQ

Seedance 2.0 FAQ

The most important constraints and workflow details for this page.

Can I use first/last frame and references together?

No. First/last frame and multimodal references are mutually exclusive scenarios, so the page blocks that combination before submit.

Which references are supported?

Reference images, videos, and audios are supported. The current page allows up to 9 images, 3 videos, and 3 audio files or public URLs. Reference video input is limited to 15 seconds total.

How are results queried?

The page creates a generation task and then reuses the existing /api/video-generations/check-result polling flow to track status until completion.

Do failed tasks refund credits?

Yes. Failed tasks continue through the site’s existing generation sync and refund path.

Can I keep results private?

Yes. Turn off public visibility to keep the task private if your account has VIP access.

Open Seedance 2.0 In One Click

Submit Seedance jobs, track progress, and download results without leaving the site workflow.