HackAfterDark ComfyUI Custom Nodes — v1.1

Some ComfyUI custom nodes that I couldn't quite find elsewhere. While I've found similar in some cases, I just haven't found them to work exactly how I'd like. I've also included some ComfyUI workflows.

They currently include;

Film/Photography aspect ratio size selector . A quick width/height select for common aspect ratios compatible with 1024x1024 resolution.

Gemini API node useful for generating images from text and/or images, generating prompts or captions from images, or enhancing prompts from text. I use this a lot for creating prompts from images I find and like. I also use it for captioning images for data sets to train LoRAs.

Note that I will continue to update these nodes and add more system prompt presets, in particular to generate prompts. The system prompt preset I have currently in here is for Z-Image Turbo and it includes instructions from the Z-Image Turbo prompting guide. It creates really good prompts from images (or text, so like a prompt enhancer) for Z-Image Turbo.

In the zip file are some ComfyUI Workflow JSON files as well. Other than those files, the contents are the same as what is found on my GitHub. You can always find the latest version of these custom nodes here: https://github.com/hackafterdark/ComfyUI-HackAfterDark-Nodes

My custom nodes should also be listed in the ComfyUI Manager as well if you wish to install from the manager. Read the README file. ComfyUI may not install dependencies upon startup, so you may need to run a pip install command. Also very important: you may be wondering where to put your Google API Key. It's in ComfyUI's settings area. I did not include a place to put an API key as input into a custom node in order to prevent accidental credential leak. The image metadata will not contain your API key. Do not use any custom nodes that have you paste API keys into the node input fields.