ComfyUI Extension: Save Image with Selective Metadata

In an effort to clean up image metadata, this custom node allows users to selectively choose what to add to the generated image's metadata. As a result, images will not be able to load ComfyUI workflows.

Currently, there are 5 inputs allowed, with each needing a key and corresponding metadata pair. If the key is empty, that key:metadata pair will be ignored.

The most prevalent use case for such selectivity is being able to save the prompt in a clear manner, as seen below.

• To do so, the easiest method is to use a String node when typing the prompt, which can output text to both CLIP Text Encode (Prompt) as well as the Save Image (Selective Metadata) node. An example of this can be seen in the below workflow.
  • (such as Multiline String which has been packaged with this node for convenience)

<img width="970" height="551" alt="image" src="https://github.com/user-attachments/assets/140c99cc-9f3f-406c-b1c7-daa69686ae19" /> <img width="2088" height="967" alt="workflow" src="https://github.com/user-attachments/assets/1ec4774f-5616-44aa-855b-7f39157dddf7" />

To install, simply follow the below instructions:

1. Navigate to ComfyUI directory in a command prompt / terminal window
2. cd custom_nodes
3. git clone https://github.com/brucew4yn3rp/ComfyUI_SelectiveMetadata/
4. cd ComfyUI_SelectiveMetadata
5. pip install -r requirements.txt

If you use a venv with ComfyUI, make sure to activate it before doing step 5. This is a pretty lightweight node so you should already have the Python libraries required (numpy, Pillow)