Skip to main content
Seedream 5.5 is ByteDance Seed’s image model, released on November 9, 2026 as the successor to Seedream 5.0 Pro. It runs in ComfyUI through the partner API, so there is nothing to download: each run is billed to your Comfy API account. Three things changed from 5.0 Pro. Unedited areas stay at the pixel level through long edit chains, so a picture survives dozens of rounds without color drift or faces sagging. The model returns real transparency instead of a fake backdrop. And text inside the image renders more reliably, which matters for posters, packaging and stickers. ByteDance says a Pro version of the 5.5 series will follow.

​
What it does

  • Text to image: one prompt in, one image out, at any preset from 1K to 2K
  • Image edit: up to 10 reference images, one output, source resolution kept when you feed the size back in
  • Transparent output: choose transparent on the node and the returned image carries an alpha channel
  • Typography: captions, labels and headline text render inside the picture, no separate overlay step
  • One image per run: batch and sequential generation are not exposed for this model
To use the Partner Nodes, you need to ensure that you are logged in properly and using a permitted network environment. Please refer to the Partner Nodes Overview section of the documentation to understand the specific requirements for using the Partner Nodes.
Make sure your ComfyUI is updated.Workflows in this guide can be found in the Workflow Templates. If you can’t find them in the template, your ComfyUI may be outdated.If nodes are missing when loading a workflow, possible reasons:
  1. You are not using the latest ComfyUI version (Nightly version)
  2. Some nodes failed to import at startup

​
Seedream 5.5 text to image

A prompt goes in and one image comes back. The template opens at (1K) 1024x1024 (1:1). Seedream 5.5 text to image workflow preview

Download Workflow

Download JSON or search “Seedream 5.5” in Template Library

​
Seedream 5.5 image edit

Load a picture, describe the change, and repeat what must stay the same. The template reads the source size with Get Image Size and feeds it back, so the result comes out at the source resolution and aspect ratio. Seedream 5.5 image edit workflow preview

Download Workflow

Download JSON or search “Seedream 5.5” in Template Library
Input materials Download the sample picture to try the edit as shipped:

api_bytedance_seedream_5_5_image_edit_input.png

Source portrait the template edits

​
Size and transparency

The size preset covers 1K, 1.5K and 2K in eight aspect ratios each, plus Custom. Custom width and height accept 240 to 8594 px, but the model only renders between 0.92 MP and 4.62 MP, so a small canvas is rejected before the request goes out. Aspect ratio must stay between 1:16 and 16:1. At the 4.62 MP ceiling a square tops out near 2150 x 2150 px, which is why the largest square preset is (2K) 2048x2048. The edit template inherits that floor: a source picture below 0.92 MP has to be scaled up before it can be edited. background has three values: auto decides from the prompt and returns transparency mainly for stickers and logos, opaque paints a backdrop, and transparent forces an alpha channel. Transparent works best when the prompt describes an isolated subject on a plain background. Save the result as PNG to keep the alpha channel; JPEG drops it.

​
Get started

  1. Update ComfyUI to 0.45.0 or later.
  2. Open the Template Library, search Seedream 5.5, and open either workflow.
  3. On the node, pick the model seedream 5.5 and a size preset, then run.
Switch the model dropdown on the node to another Seedream version to reuse the same workflow. Seedream 5.0 Pro and Flash stay at selectable presets too, while 5.0 Lite and 4.5 expose 3K and 4K rows.