❌[DEPRECATED] Style & Prompt Encoder (old version)

This is the node that started the Z-Image Power Nodes style system - the original Style & Prompt Encoder, now sitting in the pack's __deprecated folder with a ❌ on its name. If you've loaded an early Z-Image workflow, this is what you'll find where the current Style & Prompt Encoder (StylePromptEncoder2) now lives. It works, it's frozen, and the successor is strictly better.

What it does

The job is identical to the current node: take a CLIP, take your text, apply a selected style to the prompt, encode through the CLIP, and hand back CONDITIONING plus the styled prompt as a STRING. The structural difference is right in the inputs: instead of one flat style dropdown, this version has a category dropdown (photo / illustration / etc.) and a style dropdown nested under it - a two-level picker that feels clunkier than the merged 100+ style list the successor ships, and which the successor dropped.

The customization input is the same one you know: a multiline string defining custom styles in >>> blocks with {$@} as the prompt placeholder. That format has survived unchanged through every version of this pack, so anything you wrote here still works in the current encoder.

Inputs & outputs

clip (CLIP) - text encoder (Qwen3-4B for Z-Image).
category (COMBO) - the style category.
style (COMBO, default "Phone Photo") - the style within that category.
text (STRING, multiline) - base prompt.
customization (optional STRING) - custom style definitions.
Outputs: CONDITIONING + STRING (styled prompt).

Install

Ships with the pack:

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

Restart ComfyUI (or ComfyUI Manager, search "Z-Image Power Nodes"). No pip dependencies.

Use it or not?

The only reason to touch this node is a legacy workflow. The current Style & Prompt Encoder has every style this one does, a single searchable list, a visual gallery, and the author's ongoing attention. "Deprecated" here means "frozen and superseded," and unlike the pack's experimental nodes, there's no new feature to justify keeping it around. If you're porting an old workflow, swap the node, re-pick your style from the merged dropdown, and delete the category hop - the output will be at least as good, and you'll be back on the maintained path. If you're new to the pack, don't seek this one out; you'll only find it by loading old files.

Name	Type	Description
CONDITIONING	CONDITIONING	The encoded text used to guide the image generation.
STRING	STRING	The prompt after applying the selected visual style.

❌[DEPRECATED] Style & Prompt Encoder (old version)

What it does

Inputs & outputs

Install

Use it or not?

Inputs (5)

Outputs (2)

Name	Type	Default	Description
clip	CLIP		The CLIP model used for encoding the text.
category	COMBO		The category of styles you want to select from.
style	COMBO	"Phone Photo"	The style you want for your image.
text	STRING		The prompt to encode.
customizationopt	STRING		An optional multi-line string to customize existing styles. Each style definition must start with ">>>" followed by the style name, and then include its description on the next lines. The description should incorporate "{$@}" where the main text prompt will be inserted.