> ## Documentation Index
> Fetch the complete documentation index at: https://docs.jiekou.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Seedance 2.0 Video Generation

The Seedance 2.0 series models support multimodal content inputs including images, videos, audio, and text. They provide capabilities such as video generation, video editing, and video extension, and can accurately restore object details, timbre, effects, style, camera movement, and more while maintaining stable character features. Supported modes include text-to-video, image-to-video (first frame / first and last frames), and multimodal-reference video generation (image + video + audio combinations). A standard version (seedance-2.0) and a fast version (seedance-2.0-fast) are available; the fast version costs less and generates faster.

### Minimum Charge Notice

* **Applicable SKU**: Multimodal-reference video generation (with video input, i.e., the MULTI\_REF\_VID series)
* **Billing rule**: Actual charge = max(unit price per second × total video seconds, minimum charge)
* **Trigger scenario**: When the user inputs a very short video (such as 1–2 seconds) and the output is also short, the amount calculated by seconds may be lower than the cost corresponding to the supplier's minimum token consumption. In this case, the minimum charge is applied.
* **Example scenario**: A customer wants to generate a 4-second product promotional video. They upload a 2-second product promotional video and want to modify the background and colors, with no other input elements. The video is relatively simple, and the unit price per second × video seconds calculates to only $0.19 for this request. However, because a video asset was uploaded, the minimum charge is triggered, and the 4-second minimum charge ($0.30) is applied directly.

#### Minimum Charge Table

| Output Seconds | 2.0-480P | 2.0-720P | 2.0-1080P | fast-480P | fast-720P |
| -------------- | -------- | -------- | --------- | --------- | --------- |
| 4              | \$0.30   | \$0.65   | \$1.64    | \$0.23    | \$0.50    |
| 5              | \$0.39   | \$0.84   | \$2.06    | \$0.30    | \$0.64    |
| 6              | \$0.43   | \$0.93   | \$2.47    | \$0.33    | \$0.71    |
| 7              | \$0.52   | \$1.11   | \$2.88    | \$0.40    | \$0.85    |
| 8              | \$0.61   | \$1.30   | \$3.29    | \$0.46    | \$1.00    |
| 9              | \$0.65   | \$1.39   | \$3.70    | \$0.50    | \$1.07    |
| 10             | \$0.73   | \$1.58   | \$4.11    | \$0.56    | \$1.21    |
| 11             | \$0.82   | \$1.76   | \$4.52    | \$0.63    | \$1.35    |
| 12             | \$0.86   | \$1.86   | \$4.93    | \$0.66    | \$1.43    |
| 13             | \$0.95   | \$2.04   | \$5.35    | \$0.73    | \$1.57    |
| 14             | \$1.04   | \$2.23   | \$5.76    | \$0.79    | \$1.71    |
| 15             | \$1.08   | \$2.32   | \$6.17    | \$0.83    | \$1.78    |

<Tip>
  This is an **asynchronous** API and only returns the task\_id of the asynchronous task. You should use this task\_id to request the [Get Task Result API](/en/docs/models/reference-get-async-task-result) to retrieve the generated result.
</Tip>

## Request Headers

<ParamField header="Content-Type" type="string" required={true}>
  Enum value: `application/json`
</ParamField>

<ParamField header="Authorization" type="string" required={true}>
  Bearer authentication format: Bearer \{\{API Key}}.
</ParamField>

## Request Body

<ParamField body="fast" type="boolean" default={false}>
  Whether to use the fast version of the model (seedance-2.0-fast). The fast version costs less and generates faster.
</ParamField>

<ParamField body="seed" type="integer" default={-1}>
  Random seed, used to control the randomness of the generated content. Value range: \[-1, 2^32-1], where -1 means random.

  Value range: \[-1, +∞]
</ParamField>

<ParamField body="image" type="string">
  First-frame image URL or Base64 encoding. Used for image-to-video first-frame mode. Supported formats: jpeg/png/webp/bmp/tiff/gif. Aspect ratio range: (0.4, 2.5). Width and height pixel range: (300, 6000). Each image must not exceed 30MB.
</ParamField>

<ParamField body="ratio" type="string" default="adaptive">
  Aspect ratio of the generated video. adaptive means the most suitable aspect ratio is automatically selected based on the input.

  Allowed values: `16:9`, `4:3`, `1:1`, `3:4`, `9:16`, `21:9`, `adaptive`
</ParamField>

<ParamField body="prompt" type="string">
  Text prompt describing the desired video to generate. Chinese and English are supported. It is recommended that Chinese prompts do not exceed 500 characters and English prompts do not exceed 1000 words. Required in text-to-video mode; optional in other modes.
</ParamField>

<ParamField body="duration" type="integer" default={5}>
  Duration of the generated video (seconds). Range: \[4,15]

  Value range: \[4, 15]
</ParamField>

<ParamField body="watermark" type="boolean" default={false}>
  Whether the generated video contains a watermark.
</ParamField>

<ParamField body="last_image" type="string">
  Last-frame image URL or Base64 encoding. Must be provided together with the image field to implement image-to-video first-and-last-frame mode. Passing only last\_image without image is invalid. If the aspect ratios of the first-frame and last-frame images are inconsistent, the first frame takes precedence, and the last frame is automatically cropped to fit.
</ParamField>

<ParamField body="resolution" type="string" default="720p">
  Video resolution. 1080p is only supported by the standard version (fast=false).

  Allowed values: `480p`, `720p`, `1080p`
</ParamField>

<ParamField body="web_search" type="boolean" default={false}>
  Whether to enable web search. When enabled, the model independently determines whether to search internet content based on the prompt, which can improve timeliness but increases latency.
</ParamField>

<ParamField body="generate_audio" type="boolean" default={true}>
  Whether to generate audio synchronized with the visuals. When true, the model automatically generates matching vocals, sound effects, and background music based on the text and visual content.
</ParamField>

<ParamField body="reference_audios" type="string[]">
  List of reference audio files, used for multimodal-reference video generation mode. Each item is an audio URL or Base64 encoding. Formats: wav/mp3. Each audio duration: \[2,15]s. The total duration of all audio must not exceed 15s, and each file must not exceed 15MB. Audio cannot be input alone; at least 1 reference image or video must be included.

  Array length: 1 - 3
</ParamField>

<ParamField body="reference_images" type="string[]">
  List of reference images, used for multimodal-reference video generation mode. Each item is an image URL or Base64 encoding. Up to 9 images are supported. You can specify how to combine images through the prompt. The recommended format is "\[Image 1]xxx, \[Image 2]xxx".

  Array length: 1 - 9
</ParamField>

<ParamField body="reference_videos" type="string[]">
  List of reference videos, used for multimodal-reference video generation mode. Each item is a video URL. Formats: mp4/mov. Resolution: 480p/720p. Each video duration: \[2,15]s. The total duration of all videos must not exceed 15s, and each file must not exceed 50MB.

  Array length: 1 - 3
</ParamField>

<ParamField body="return_last_frame" type="boolean" default={false}>
  Whether to return the last-frame image of the generated video (png format, no watermark). This can be used for continuous video generation: use the last frame as the first frame of the next video segment.
</ParamField>

## Response Information

<ResponseField name="task_id" type="string" required={true}>
  Use task\_id to request the [Get Task Result API](/en/docs/models/reference-get-async-task-result) to retrieve the generated output.
</ResponseField>