ByteDanceSeedreamNodeV3 - ComfyUI Built-in Node Documentation

ByteDance Seedream 4.5 & 5.0 generates images from a text prompt (text-to-image) or generates/edits images guided by optional reference images, using ByteDance Seedream 4.0, 4.5, and 5.0 models at up to 4K resolution. The node sends the prompt and any reference images to the ByteDance API, waits for the generation task to complete, and returns the resulting image tensor or tensors.

Inputs

Common Inputs

Parameter	Description	Data Type	Required	Range
`prompt`	Text prompt for creating or editing an image. Must not be empty after trimming whitespace.	STRING	Yes	Multiline text
`model`	Selects the Seedream model to use. Each model exposes its own set of sub-parameters and limits below.	DYNAMIC_COMBO	Yes	”seedream 5.0 pro" "seedream 5.0 lite" "seedream-4-5-251128" "seedream-4-0-250828”

Seedream 5.0 Pro Inputs (seedream 5.0 pro)

Parameter	Description	Data Type	Required	Range
`size_preset`	Pick a recommended size. Select Custom to use the width and height below. Default: first recommended preset for this model.	COMBO	No	Model-specific recommended size presets ”Custom”
`width`	Custom width for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 3136 (step 2)
`height`	Custom height for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 2496 (step 2)
`prompt_optimization`	Prompt-optimization mode when reference images are provided: ‘standard’ gives higher quality, ‘fast’ shorter generation time. Default: “standard”.	COMBO	No	”standard" "fast”
`seed`	Seed to use for generation. Default: 42.	INT	No	0 to 2147483647
`watermark`	Whether to add an “AI generated” watermark to the image. Default: false.	BOOLEAN	No	true / false
`thinking`	Enable the model’s prompt-optimization reasoning (‘thinking’) for better adherence. Can substantially increase generation time — notably on Seedream 5.0 Pro. Can only be disabled for text-to-image (not when reference images are provided). Default: true.	BOOLEAN	No	true / false

Seedream 5.0 Lite Inputs (seedream 5.0 lite)

Parameter	Description	Data Type	Required	Range
`size_preset`	Pick a recommended size. Select Custom to use the width and height below. Default: first recommended preset for this model.	COMBO	No	Model-specific recommended size presets ”Custom”
`width`	Custom width for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 6240 (step 2)
`height`	Custom height for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 4992 (step 2)
`max_images`	Maximum number of images to generate. With 1, exactly one image is produced. With >1, the model generates between 1 and max_images related images (e.g., story scenes, character variations). Total images (input + generated) cannot exceed 15. Default: 1.	INT	No	1 to 14
`fail_on_partial`	If enabled, abort execution if any requested images are missing or return an error. Default: false.	BOOLEAN	No	true / false
`seed`	Seed to use for generation. Default: 42.	INT	No	0 to 2147483647
`watermark`	Whether to add an “AI generated” watermark to the image. Default: false.	BOOLEAN	No	true / false
`thinking`	Enable the model’s prompt-optimization reasoning (‘thinking’) for better adherence. Can substantially increase generation time — notably on Seedream 5.0 Pro. Can only be disabled for text-to-image (not when reference images are provided). Default: true.	BOOLEAN	No	true / false

Seedream 4.5 Inputs (seedream-4-5-251128)

Parameter	Description	Data Type	Required	Range
`size_preset`	Pick a recommended size. Select Custom to use the width and height below. Default: first recommended preset for this model.	COMBO	No	Model-specific recommended size presets ”Custom”
`width`	Custom width for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 6240 (step 2)
`height`	Custom height for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 4992 (step 2)
`max_images`	Maximum number of images to generate. With 1, exactly one image is produced. With >1, the model generates between 1 and max_images related images (e.g., story scenes, character variations). Total images (input + generated) cannot exceed 15. Default: 1.	INT	No	1 to 10
`fail_on_partial`	If enabled, abort execution if any requested images are missing or return an error. Default: false.	BOOLEAN	No	true / false
`seed`	Seed to use for generation. Default: 42.	INT	No	0 to 2147483647
`watermark`	Whether to add an “AI generated” watermark to the image. Default: false.	BOOLEAN	No	true / false
`thinking`	Enable the model’s prompt-optimization reasoning (‘thinking’) for better adherence. Can substantially increase generation time — notably on Seedream 5.0 Pro. Can only be disabled for text-to-image (not when reference images are provided). Default: true.	BOOLEAN	No	true / false

Seedream 4.0 Inputs (seedream-4-0-250828)

Parameter	Description	Data Type	Required	Range
`size_preset`	Pick a recommended size. Select Custom to use the width and height below. Default: first recommended preset for this model.	COMBO	No	Model-specific recommended size presets ”Custom”
`width`	Custom width for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 6240 (step 2)
`height`	Custom height for image. Value is working only if `size_preset` is set to `Custom`. Default: 2048.	INT	No	1024 to 4992 (step 2)
`max_images`	Maximum number of images to generate. With 1, exactly one image is produced. With >1, the model generates between 1 and max_images related images (e.g., story scenes, character variations). Total images (input + generated) cannot exceed 15. Default: 1.	INT	No	1 to 10
`fail_on_partial`	If enabled, abort execution if any requested images are missing or return an error. Default: false.	BOOLEAN	No	true / false
`seed`	Seed to use for generation. Default: 42.	INT	No	0 to 2147483647
`watermark`	Whether to add an “AI generated” watermark to the image. Default: false.	BOOLEAN	No	true / false
`thinking`	Enable the model’s prompt-optimization reasoning (‘thinking’) for better adherence. Can substantially increase generation time — notably on Seedream 5.0 Pro. Can only be disabled for text-to-image (not when reference images are provided). Default: true.	BOOLEAN	No	true / false

Reference Inputs

Parameter	Description	Data Type	Required	Range
`images`	Growable slot: optional reference image(s) for image-to-image or multi-reference generation. Connect 1..N images (e.g., `image_1`, `image_2`, …); the count limit is per model (see notes below). If a connected image contains a batch of images, every image in the batch counts toward the limit.	IMAGE	No	0 to 10 (Seedream 5.0 Pro, Seedream 4.5, Seedream 4.0) 0 to 14 (Seedream 5.0 Lite)

Notes:

The prompt must not be empty after trimming whitespace.
Maximum number of reference images: 10 for Seedream 5.0 Pro, Seedream 4.5, and Seedream 4.0; 14 for Seedream 5.0 Lite.
Each reference image must have an aspect ratio between 1:3 and 3:1.
When max_images is greater than 1 (not available on Seedream 5.0 Pro), the total number of reference images plus generated images cannot exceed 15.
thinking can only be disabled for text-to-image generation. When reference images are provided, thinking must be enabled.
width and height are only used when size_preset is set to “Custom”.
prompt_optimization is only available on Seedream 5.0 Pro.
max_images and fail_on_partial are only available on Seedream 5.0 Lite, Seedream 4.5, and Seedream 4.0; Seedream 5.0 Pro always requests a single image.
Resolution requirements (width x height):
- Seedream 5.0 Pro: between 0.92MP (921,600 pixels) and 4.19MP (4,194,304 pixels).
- Seedream 5.0 Lite and Seedream 4.5: at least 3.68MP (3,686,400 pixels).
- Seedream 4.0: at least 0.92MP (921,600 pixels).
- All non-Pro models: at most 16.78MP (16,777,216 pixels).

Outputs

Output Name	Description	Data Type
`image`	The generated image tensor. When multiple images are generated, they are concatenated into a single batched IMAGE tensor.	IMAGE

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): 68dd23afdb5720491cef784b22ad66ff0baf80984ea652ea4c13e6c264c029ac

​Inputs

​Common Inputs

​Seedream 5.0 Pro Inputs (seedream 5.0 pro)

​Seedream 5.0 Lite Inputs (seedream 5.0 lite)

​Seedream 4.5 Inputs (seedream-4-5-251128)

​Seedream 4.0 Inputs (seedream-4-0-250828)

​Reference Inputs

​Outputs