Seedance 2.0 Pro

Video

High-fidelity video · up to 4k

What Is Seedance 2.0 API?

Overview

Seedance 2.0 is ByteDance Seed's multimodal AI video generation model built on a unified audio-video architecture. It supports text, image, video, and audio inputs for video generation, reference-based creation, editing, and extension.

Available workflows:

Text to Video
Image to Video
First-Last Frame to Video
Multi-Reference to Video

Key Capabilities

15-second audio-video generation: Generate multi-shot videos up to 15 seconds with synchronized audio.
Multimodal references: Combine up to 9 images, 3 videos, and 3 audio files in one generation.
Video editing and extension: Modify existing content or continue a video with natural-language instructions.
Native synchronized audio: Generate dialogue, sound effects, ambience, and background music with the video.
Complex motion and camera control: Improved instruction following for actions, interactions, composition, and camera movement.

Inputs

Seedance 2.0 supports up to 12 reference files, with a maximum of 9 images, 3 videos, and 3 audio files.

Modality	Limit	Formats	Constraints
Images	Up to 9	JPEG, PNG, WebP, GIF, BMP, TIFF, HEIC	Max 30 MB each; shortest side ≥ 400 px
Videos	Up to 3	MP4, MOV	Max 50 MB each; up to 15s each; must be uploaded as assets
Audio	Up to 3	MP3, WAV	Max 15 MB each; up to 15s each; requires at least one image or video

The total number of reference assets across images, videos, and audio must not exceed 12. `audioUrls` cannot be the only reference — it requires at least one image or video.

Parameters

These fields follow the current sd2 API schema in SeeGen AI. resolution controls the aspect ratio/canvas, while outputResolution controls the final resolution tier.

Parameter	Type	Default	Options / Description
prompt	string	—	Text description of the video. Max 20,000 characters.
urls	string[]	—	Reference images. Supports HTTP URLs and asset:// references. Up to 9 images in reference mode.
videoUrls	string[]	—	Reference video asset:// URIs. Up to 3 videos; external video URLs are not supported.
audioUrls	string[]	—	Reference audio inputs. Up to 3 audio files; audio cannot be used alone and requires at least one image or video.
videoInputMode	string	"keyframe"	"keyframe" for image / first-last frame generation; "reference" for multimodal reference generation.
mode	string	—	Omit for standard reference generation; use "edit" for video editing or "extend" for video continuation.
duration	string	"5s"	Output duration from "4s" to "15s".
resolution	string	"auto"	Aspect ratio: auto, 720x720, 720x960, 960x720, 1280x720, 720x1280, 1280x540.
outputResolution	string	"720p"	480p, 720p, 1080p, 2k, 4k. Seedance 2.0 Pro outputs 480p, 720p, 1080p, and 4K natively; 2K is automatically upscaled by SeeGen AI.
generateAudio	boolean	true	Generate synchronized speech, sound effects, and background music. Set false for silent video.
bitrateMode	string	"standard"	"standard" or "high". High bitrate preserves more detail at a larger file size; pricing is unchanged.
seed	integer	-1	-1 or omit for a random seed; 0–2147483647 for a fixed seed.

Seedance 2.0 Pro outputs 480p, 720p, 1080p, and 4K natively. 2K output is automatically upscaled.

Seedance 2.0 API Pricing

Seedance 2.0 uses pay-as-you-go credit pricing. The rates below are based on the $500 API Pack: 125,000 credits at $0.004 per credit.

Condition	Rate	10-sec clip
SD2-Pro 480p, no video input	20 credits/sec (~$0.08/sec)	~200 credits (~$0.80)
SD2-Pro 480p, with video input	15 credits/sec (~$0.06/sec)	Based on output + reference video duration
SD2-Pro 720p, no video input	40 credits/sec (~$0.16/sec)	~400 credits (~$1.60)
SD2-Pro 720p, with video input	30 credits/sec (~$0.12/sec)	Based on output + reference video duration
SD2-Pro 1080p (native), no video input	100 credits/sec (~$0.40/sec)	~1,000 credits (~$4.00)
SD2-Pro 1080p (native), with video input	75 credits/sec (~$0.30/sec)	Based on output + reference video duration
SD2-Pro 2K (upscaled), no video input	70 credits/sec (~$0.28/sec)	~700 credits (~$2.80)
SD2-Pro 2K (upscaled), with video input	30 × (out + in) + 30 × out	Based on output + reference video duration
SD2-Pro 4K (native), no video input	200 credits/sec (~$0.80/sec)	~2,000 credits (~$8.00)
SD2-Pro 4K (native), with video input	150 credits/sec (~$0.60/sec)	Based on output + reference video duration

Credit Calculation

1. No video input

credits = out × rate

2. With video input

credits = (out + in) × rate

out = output_seconds. in = the sum of ⌈duration⌉ for each input video. Each duration is rounded up, with a minimum input duration of ceil(out × 2 / 3). For 2K upscale, the additional upscale rate applies only to the output duration.

Run Seedance 2.0 API

Generate a video in two steps: create a task, then query the result using the returned taskId.

Asset Requirements

Upload reference images, videos, and audio files as assets first. Once their status becomes ACTIVE, use the returned asset:// URIs in your generation request.

Images → urls
Videos → videoUrls
Audio → audioUrls

All reference assets must be reviewed and ACTIVE before use.

View Asset Upload Guide →

Create a Task

curl -X POST https://seegen.ai/api/v1/jobs/createTask \
  -H "Authorization: Bearer $SEEGEN_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "sd2",
    "inputs": {
      "urls": [
        "asset://asset-reference-image"
      ],
      "videoUrls": [
        "asset://asset-reference-video"
      ],
      "audioUrls": [
        "asset://asset-reference-audio"
      ],
      "prompt": "Create a cinematic scene using the provided character, motion, and audio references.",
      "duration": "5s",
      "videoInputMode": "reference",
      "resolution": "1280x720",
      "outputResolution": "720p",
      "generateAudio": true
    }
  }'

The request structure above matches SeeGen's current Multi-Reference API schema for Seedance 2.0.

Query the Result

Use the returned taskId to check the task status. When the status becomes COMPLETED, the generated video is available in output[0].url.

curl \
  -H "Authorization: Bearer $SEEGEN_API_KEY" \
  "https://seegen.ai/api/v1/jobs/queryTask?taskId=YOUR_TASK_ID"

A completed task returns status: "COMPLETED" and the generated video URL in output[0].url.

Seedance 2.0 API Use Cases

Use Case	15-Second Generation	Multimodal References	Video Editing	Native Audio
Film & Entertainment	Multi-shot scenes and short narrative sequences	Combine characters, environments, motion, and style references	Extend scenes or modify characters and actions	Generate dialogue, music, ambience, and SFX
Advertising & E-commerce	Product ads, brand stories, and social videos	Maintain products, models, brand identity, and visual style	Replace products, labels, models, or scene details	Add voiceovers, product sounds, and music
Education & Knowledge	Short explainers and visual demonstrations	Combine diagrams, characters, scenes, and reference footage	Refine presenters, demonstrations, or visual elements	Generate narration and synchronized sound
Industrial & Enterprise	Workflow, training, and product demonstration videos	Maintain consistent equipment, products, and environments	Update labels, specifications, or product variants	Add instructional narration and sound effects
Games & Virtual Content	CG trailers, character scenes, and episodic content	Combine virtual characters, environments, and motion references	Modify characters, styles, and visual details	Generate dialogue, music, and game-style SFX

Seedance 2.0 Pro vs Seedance 2.0 Fast

Feature	Seedance 2.0 Pro	Seedance 2.0 Fast
Model ID	sd2	sd2-fast
Maximum duration	Up to 15 seconds	Up to 15 seconds
Reference inputs	Up to 9 images + 3 videos + 3 audio	Up to 9 images + 3 videos + 3 audio
Native resolution	480p, 720p, 1080p, 4K	480p, 720p
Higher resolutions	2K upscaled	1080p, 2K, 4K upscaled
480p, no video input	20 credits/sec	16 credits/sec
720p, no video input	40 credits/sec	32 credits/sec
5s 720p example	200 credits (~$0.80)	160 credits (~$0.64)
Best for	Native high-resolution output	Lower-cost generation and iteration

Choose Seedance 2.0 Pro when you need native 1080p or 4K output. Choose Seedance 2.0 Fast when lower generation cost is more important.

FAQs about Seedance 2.0 API

What is Seedance 2.0 API?

Seedance 2.0 API provides programmatic access to ByteDance Seed's multimodal video generation model, supporting text, image, video, and audio inputs for generation and reference-based workflows.

Is Seedance 2.0 API available on SeeGen AI?

Yes. SeeGen AI provides Seedance 2.0 Pro as sd2, Seedance 2.0 Fast as sd2-fast, and Seedance 2.0 Mini as sd2-mini.

How long can Seedance 2.0 videos be?

Seedance 2.0 supports output durations from 4 to 15 seconds, with 5 seconds as the default API duration.

What reference inputs does Seedance 2.0 support?

A Seedance 2.0 multi-reference task can use up to 9 images, 3 videos, and 3 audio files, with a maximum of 12 reference files in total.

Can audio be used as the only reference?

No. In the current SeeGen Seedance 2.0 API, audioUrls requires at least one image or video reference.

What output resolutions does Seedance 2.0 Pro support?

SeeGen AI supports 480p, 720p, 1080p, 2k, and 4k. Seedance 2.0 Pro outputs 480p, 720p, 1080p, and 4K natively, while 2K is automatically upscaled.

What is the difference between Seedance 2.0 Pro and Fast?

Both support the same major Seedance 2.0 generation workflows. Pro supports native 1080p and 4K output, while Fast generates natively at 480p/720p and uses automatic upscaling for higher resolutions. Fast also uses fewer credits at 480p and 720p.

Does Seedance 2.0 API generate audio?

Yes. generateAudio defaults to true and generates synchronized speech, sound effects, and background music. Set it to false if you need silent output.