A node is one block on the ComfyUI canvas. Each node here calls a BytePlus service directly, billed to your own BytePlus account, with a key you save once in Settings → BytePlus (see API keys). Generation nodes only run when something uses their output, such as Save Image or Save Video, and one with nothing connected is not run and not billed. The exceptions are the two lookup nodes, Video Query Tasks and Asset Library, which run on their own.
All 22 nodes are listed below. Open a node’s page for its inputs, outputs, limits and common errors.
| Node |
What it does |
Key |
| Seedream 4.5 & 5.0 |
Creates or edits images from a text prompt and optional reference images, up to 4K. |
ModelArk |
| Seedream 5.0 Layer Separation |
Splits one image into a base image and up to 16 transparent layers, each with a name and position. |
ModelArk |
The first three nodes offer Seedance 2.5, 2.5 Premium and the 2.0 models in one model choice. Draft to Final has no model choice: it reads the model from the draft.
These nodes stay in the pack, but no template uses them.
| Node |
What it does |
Key |
| Video Query Tasks |
Lists your recent Seedance tasks, or looks up task IDs, as JSON so you can find a task again. |
ModelArk |
| Node |
What it does |
Key |
| BytePlus LLM |
Answers in text with Seed 2.0, Seed 2.1 Turbo, DeepSeek V4.1 Flash or GLM 5.3 Flash. It can read images and videos, and audio on Seed 2.0 Lite and Mini. |
ModelArk |
| Node |
What it does |
Key |
| Seed Audio 1.0 |
Makes speech, music, sound effects and multi-speaker dialogue from one prompt, up to 2 minutes. |
Seed Speech |
| Seed Speech TTS |
Turns text into speech with TTS 2.0 voices, TTS 1.0 speakers or cloned voices. |
Seed Speech |
| Seed Speech ASR |
Turns speech into text with timings and SRT subtitles. |
Seed Speech |
| Seed Voice Clone |
Clones a voice from a reference clip and outputs its speaker ID for TTS or Seed Audio. |
Seed Speech |
| Node |
What it does |
Key |
| vCube Video Enhance |
Upscales and restores a video, up to 8K, with an optional before and after comparison. |
AI MediaKit |
| Video Smoothness Enhance |
Repairs stutter and repeated frames without changing the resolution. |
AI MediaKit |
| Image Quality Enhance |
Upscales and restores images: sharpening, denoising and portrait, text and colour enhancement. |
AI MediaKit |
| Node |
What it does |
Key |
| Hyper3D Gen2 |
Generates a textured 3D model from a text prompt, up to five images, or both. |
ModelArk |
| Hitem3D 2.0 |
Generates a 3D model from one to four images of an object, with or without texture. |
ModelArk |
The asset nodes need Advanced Creation Rights on your BytePlus account.
| Node |
What it does |
Key |
| Create Image Asset |
Adds an image to your private asset library and outputs its asset ID. |
ModelArk, IAM access key and secret key |
| Create Video Asset |
Adds a video to your private asset library and outputs its asset ID. |
ModelArk, IAM access key and secret key |
| Create Audio Asset |
Adds an audio clip (WAV or MP3, 2 to 30 seconds) to your private asset library. |
ModelArk, IAM access key and secret key |
| Asset Library |
Lists the assets in your private library as asset:// links for Seedance references. |
ModelArk, IAM access key and secret key |
