⚡| My Top-10 Styles

My Top-10 Styles is the one-click hotbar for the Z-Image Power Nodes style system. Where the full Style & Prompt Encoder makes you open a dropdown or gallery, this node turns your ten most-used styles into ten boolean toggles: tick one, and the style template gets folded into your prompt text. It doesn't encode anything itself - it's a prompt-string transformer that hands the styled text off to whatever encoder you want, which makes it the most flexible member of the style family.

It's the consumption half of a pair. My Top-10 Styles (Editor) builds the ten-style list; this node consumes it. Wire the Editor's TOP_STYLES output into this node's top_styles input and your toggles light up with names.

How it works

You feed it your base input text, optionally feed it the top_styles config, and tick one (and only one) of style_1 through style_10. The node finds the matching style definition, strips any previously-applied "Custom" style tag from your text, and appends the new style's template. The output STRING is your prompt with the style baked in - exactly what you'd paste into a Style & Prompt Encoder's text field or a plain CLIP Text Encode.

The output_to input (options custom_1 through custom_4) is the clever part: each channel is a separate "slot," so you can chain multiple Top-10 nodes and have each one own its own slice of the prompt. Node A writes to custom_1, node B to custom_2, and the last node in the chain passes along the fully combined result.

The inputs that matter

• style_1 … style_10 (BOOL) - the toggles. Only one is meant to be on at a time.
• top_styles (optional TOP_STYLES) - from the Editor node; defines what each toggle is.
• input (optional STRING) - the prompt to modify; chain the previous Top-10 node's output here.
• output_to (COMBO) - which channel this node writes to.
• control_after_generate (fixed/next) - ComfyUI's standard widget behavior for toggles after a run.
• Output: output (STRING) - the styled prompt.

Install

cd ComfyUI/custom_nodes
git clone https://github.com/martin-rizzo/ComfyUI-ZImagePowerNodes

Restart ComfyUI (or ComfyUI Manager → "Z-Image Power Nodes"). No pip dependencies; the styles ship with the pack.

The honest take

This node shines for iteration speed - once your top ten is set, a prompt edit is literally one toggle click plus a queue, and you stop paging through a gallery mid-flow. Two things to keep in mind. First, "output_to" is per-node channeling, not a style picker: if you tick two styles in the same node, it only applies the first one found, so layer via chained nodes instead. Second, unlike the encoders, this gives you text, not conditioning - you still need a clip input somewhere down the line, which is either a feature (full control) or a footgun (easy to forget), depending on your workflow. If your whole style life is "photoreal, cinematic, or anime," this node is probably overkill - but if you juggle ten looks, it's the best part of the pack.