ByteDanceSeedNode - ComfyUI Built-in Node Documentation - ComfyUI

Generate text responses with ByteDance’s Seed 2.0 models. Provide a text prompt and optionally connect images or videos to give the model extra context. The model is chosen from the available Seed 2.0 variants, and the node returns the model’s text answer.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`prompt`	Text input to the model. (default: "")	STRING	Yes	N/A
`model`	The Seed model used to generate the response. This selector also exposes the model’s sub-parameters.	DYNAMIC_COMBO	Yes	`"Seed 2.0 Pro"` `"Seed 2.0 Lite"` `"Seed 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
`system_prompt`	Foundational instructions that dictate the model’s behavior. (default: "")	STRING	No	N/A

Model Inputs (Shared by Seed 2.0 Pro, Seed 2.0 Lite, and Seed 2.0 Mini)

All three Seed models expose the same sub-parameters when selected.

Parameter	Description	Data Type	Required	Range
`temperature`	Controls randomness. 0.0 is deterministic, higher values are more random. (default: 1.0)	FLOAT	Yes	0.0 to 2.0 (step: 0.01)

Reference Inputs

Parameter	Description	Data Type	Required	Range
`images`	Optional image(s) to use as context for the model. Up to 20 images. Growable slot: connect 1..20 items, e.g. `image_1` through `image_20`.	IMAGE	No	0 to 20 images
`videos`	Optional video(s) to use as context for the model. Up to 4 videos. Growable slot: connect 1..4 items, e.g. `video_1` through `video_4`.	VIDEO	No	0 to 4 videos

Note: The model parameter is a dynamic combo that reveals the reference and temperature sub-parameters once a model is selected. The images and videos slots are growable, so you can connect several inputs for multimodal context.

prompt is required and must contain at least one non-whitespace character; otherwise an error is raised.
A maximum of 20 images is supported per request. This limit counts all images across connected batches.
A maximum of 4 videos is supported per request.
An error is raised if the model returns an empty response, or if the model refuses to respond (the refusal text is reported).

Outputs

Output Name	Description	Data Type
`output`	The generated text response from the Seed model.	STRING

This documentation was AI-generated. If you find any errors or have suggestions for improvement, please feel free to contribute! Edit on GitHub

Source fingerprint (SHA-256): 23c9b0e9983a65ce859e2e92acfe71604297f16d711fa094a6617a9915a46020