PromptSimplification ♾️Mixlab

Prompts have a way of growing over time - you add a tag to fix one problem, another to fix a side effect of the first, and six months later you've got a wall of comma-separated keywords and no idea which ones are actually doing anything. PromptSimplification is Mixlab's answer to that: hand it a long prompt and it hands back a leaner version, or a few candidate leaner versions, so you can figure out what you can actually cut.

The README pairs this node conceptually with a sibling called PromptImage, describing the two together as helping "simplify prompt words, comparing images and prompt word nodes" - the intended workflow is iterative: generate with your full prompt, generate with a simplified candidate, compare the two images side by side, and see how much of your original prompt was actually load-bearing versus just noise you'd copied from somewhere and never questioned.

The inputs are small: prompt, a multiline STRING where you paste the prompt you want trimmed, and length (an integer, 1 to 100, default 5). The node's own schema doesn't spell out exactly what length measures - most likely it's steering how aggressively the prompt gets cut down, though whether that's word-count-per-result or something else isn't documented anywhere the pack exposes. What is clear from the schema is the output shape: prompts, plural, a STRING list rather than one string - so this isn't handing you one definitive "correct" simplified prompt, it's handing you options to compare, which lines up with the README's framing of pairing this with an image-comparison step rather than trusting a single automated answer.

Given that ambiguity, the practical approach is to just try it on a prompt you know well: run it, look at what came back, and adjust length up or down until the output style matches what you're actually after. Treat the first result as a starting point to eyeball rather than a black-box answer to trust blind - which, honestly, is decent advice for any prompt-editing tool, automated or not.

Install follows the pack's usual route. Through ComfyUI Manager: search "comfyui-mixlab-nodes," install, restart. Manually:

cd ComfyUI/custom_nodes
git clone https://github.com/shadowcz007/comfyui-mixlab-nodes.git
cd comfyui-mixlab-nodes
install.bat

or pip3 install -r requirements.txt in a venv, or the embedded-python pip call on a portable build. Restart ComfyUI once it's done.

One thing worth double-checking if the output looks unhelpful: since this node lives in the same Prompt category as PromptGenerate_Mix, and that node depends on separate downloaded models to function, it's worth confirming PromptSimplification itself doesn't have a similar dependency you're missing - the README doesn't call out a specific model download tied to this node the way it does for PromptGenerate_Mix, but if you get an outright error rather than just an unsatisfying result, check the ComfyUI console output for a missing-file message before assuming the node is simply not useful for your prompt.