> ## Documentation Index
> Fetch the complete documentation index at: https://docs.comfy.org/llms.txt
> Use this file to discover all available pages before exploring further.
# ByteDance2ReferenceNodeV2 - ComfyUI Built-in Node Documentation
> ByteDance Seedance 2.5 Reference to Video generates, edits, or extends videos using ByteDance Seedance models (Seedance 2.5, 2.0, 2.0 Fast, and 2.0 Mini) guided by a text prompt an
ByteDance Seedance 2.5 Reference to Video generates, edits, or extends videos using ByteDance Seedance models (Seedance 2.5, 2.0, 2.0 Fast, and 2.0 Mini) guided by a text prompt and optional reference images, videos, audio, or previously uploaded library assets. It uploads the references, submits a generation task, waits for completion, and returns the finished video file.
## Inputs
### Common Inputs
| Parameter | Description | Data Type | Required | Range |
| ----------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | -------------- | -------- | ------------------------------------------------------------------------------------ |
| `model` | Model selector. Seedance 2.5 for the newest model, videos up to 30 seconds and mp4/mov output; Seedance 2.0 for maximum quality and 4k; Fast for speed optimization; Mini for the fastest, lowest-cost generation. Selecting a model changes the input widgets shown below. | DYNAMIC\_COMBO | Yes | "Seedance 2.5"
"Seedance 2.0"
"Seedance 2.0 Fast"
"Seedance 2.0 Mini" |
| `seed` | Seed controls whether the node should re-run; results are non-deterministic regardless of seed. Default: 0. | INT | Yes | 0 to 2147483647 |
| `watermark` | Whether to add a watermark to the video. Default: False. Advanced setting. | BOOLEAN | Yes | true
false |
### Seedance 2.5 Inputs
These inputs appear when `model` is set to "Seedance 2.5".
| Parameter | Description | Data Type | Required | Range |
| ---------------- | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------- | -------- | ------------------------------------------------------------------------------- |
| `prompt` | Text prompt for video generation. Put spoken lines in double quotes to steer the generated dialogue. Default: empty string. | STRING | Yes | Multiline text |
| `resolution` | Resolution of the output video. Default: 720p. | COMBO | Yes | "480p"
"720p"
"1080p" |
| `ratio` | Aspect ratio of the output video. Default: 16:9. | COMBO | Yes | "16:9"
"4:3"
"1:1"
"3:4"
"9:16"
"21:9"
"adaptive" |
| `duration` | Duration of the output video in seconds (4-30). Default: 5. | INT | Yes | 4 to 30 |
| `generate_audio` | Enable audio generation for the output video. Default: True. | BOOLEAN | Yes | true
false |
| `task_type` | What to do with the reference media. Every value except auto is validated when the task is submitted, so mismatched settings fail before generation starts.
auto: the model infers the task from the prompt and inputs, and settings that conflict with its reading fail only after generation has started.
reference: generate a new video guided by the reference images, videos, and audio.
edit: change a connected reference video (add, remove, replace); the output keeps the source clip's own length and aspect ratio, and the duration and ratio widgets are ignored.
extend: continue a connected reference video forward or backward; the prompt should say "extend forward", "extend backward", or "continue", the aspect ratio follows the source clip, and the output contains only the newly generated segment of the duration you set, not the source clip. Default: auto. | COMBO | Yes | "auto"
"reference"
"edit"
"extend" |
| `output_format` | Container format of the output video. Default: mp4. | COMBO | Yes | "mp4" |
### Seedance 2.0 Inputs
These inputs appear when `model` is set to "Seedance 2.0".
| Parameter | Description | Data Type | Required | Range |
| ---------------- | ------------------------------------------------------------ | --------- | -------- | ------------------------------------------------------------------------------- |
| `prompt` | Text prompt for video generation. Default: empty string. | STRING | Yes | Multiline text |
| `resolution` | Resolution of the output video. | COMBO | Yes | "480p"
"720p"
"1080p"
"4k" |
| `ratio` | Aspect ratio of the output video. Default: adaptive. | COMBO | Yes | "16:9"
"4:3"
"1:1"
"3:4"
"9:16"
"21:9"
"adaptive" |
| `duration` | Duration of the output video in seconds (4-15). Default: 7. | INT | Yes | 4 to 15 |
| `generate_audio` | Enable audio generation for the output video. Default: True. | BOOLEAN | Yes | true
false |
### Seedance 2.0 Fast and Seedance 2.0 Mini Inputs
These inputs appear when `model` is set to "Seedance 2.0 Fast" or "Seedance 2.0 Mini". Both models share the same input set.
| Parameter | Description | Data Type | Required | Range |
| ---------------- | ------------------------------------------------------------ | --------- | -------- | ------------------------------------------------------------------------------- |
| `prompt` | Text prompt for video generation. Default: empty string. | STRING | Yes | Multiline text |
| `resolution` | Resolution of the output video. | COMBO | Yes | "480p"
"720p" |
| `ratio` | Aspect ratio of the output video. Default: adaptive. | COMBO | Yes | "16:9"
"4:3"
"1:1"
"3:4"
"9:16"
"21:9"
"adaptive" |
| `duration` | Duration of the output video in seconds (4-15). Default: 7. | INT | Yes | 4 to 15 |
| `generate_audio` | Enable audio generation for the output video. Default: True. | BOOLEAN | Yes | true
false |
### Reference Inputs
These growable reference slots are available for all models. The maximum number of slots differs per model: Seedance 2.5 supports up to 30 images, 10 videos, 10 audios, and 30 assets; Seedance 2.0, 2.0 Fast, and 2.0 Mini support up to 9 images, 3 videos, 3 audios, and 9 assets.
| Parameter | Description | Data Type | Required | Range |
| ------------------ | --------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------- | --------- | -------- | ---------------------------------------------------------------- |
| `reference_images` | Growable slot: connect 1..N reference images that guide the output. The count limit is per model (see model sections). Images are validated for aspect ratio (0.4 to 2.5) and auto-downscaled to a maximum side of 6000 pixels. | IMAGE | No | 1..9 slots (Seedance 2.0 family)
1..30 slots (Seedance 2.5) |
| `reference_videos` | Growable slot: connect 1..N reference videos. The count limit is per model (see model sections). Each video must be at least 1.8 seconds long and must fit the pixel limits for the selected model and resolution. | VIDEO | No | 1..3 slots (Seedance 2.0 family)
1..10 slots (Seedance 2.5) |
| `reference_audios` | Growable slot: connect 1..N reference audio tracks. The count limit is per model (see model sections). Each audio must be at least 1.8 seconds long. | AUDIO | No | 1..3 slots (Seedance 2.0 family)
1..10 slots (Seedance 2.5) |
| `reference_assets` | Growable slot: connect 1..N asset ID strings for media already uploaded to the Seedance virtual library. Each asset must be Active. You can refer to an asset in the prompt with tokens such as `asset1` or `asset 1`; the node replaces them with the asset's positional label (for example "Image 2" or "Video 1"). | STRING | No | 1..9 slots (Seedance 2.0 family)
1..30 slots (Seedance 2.5) |
| `auto_downscale` | Automatically downscale reference videos that exceed the model's pixel budget for the selected resolution. Aspect ratio is preserved; videos already within limits are untouched. Default: True. | BOOLEAN | No | true
false |
| `auto_upscale` | Automatically upscale reference videos that are below the model's minimum pixel count for the selected resolution. Aspect ratio is preserved; videos already meeting the minimum are untouched. Note: upscaling a low-resolution source does not add real detail and may produce lower-quality generations. Default: False. Advanced setting. | BOOLEAN | No | true
false |
**Note:** At least one reference image, video, or asset is required to run the node (Seedance 2.5 also accepts audio-only references). Reference videos and audios must each be at least 1.8 seconds long, and the combined duration of all reference videos (and separately, all reference audios) must not exceed the selected model's maximum total seconds. Reference images must have an aspect ratio between roughly 2:5 and 5:2 (0.4 to 2.5), be at least 300x300 pixels, and are automatically downscaled to a maximum side of 6000 pixels. The `task_type` "edit" and "extend" options are only available with Seedance 2.5 and both require at least one reference video; when "edit" is used, the output keeps the source clip's own length and aspect ratio and the `duration` and `ratio` widgets are ignored, and when "extend" is used, the output contains only the newly generated segment at the duration you set. Referenced assets must be in Active status, otherwise the task fails.
## Outputs
| Output Name | Description | Data Type |
| ----------- | -------------------------------------------------------------------------------------------------------------------------------------- | --------- |
| `video` | The generated video, downloaded from the provider once the generation task completes. Contains audio when audio generation is enabled. | VIDEO |
> This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! [Edit on GitHub](https://github.com/Comfy-Org/embedded-docs/blob/main/comfyui_embedded_docs/docs/ByteDance2ReferenceNodeV2/en.md)
***
**Source fingerprint (SHA-256):** `3a6bba12e719204ba5dba9d7d5f2b4c5285ed68974ee015b6e4a7892a1cf0933`