ComfyUI Node

📷Prompt_Styler

PhotoMaker's style templates, as a standalone node

By ZHO-ZHO-ZHO·Created 3 years ago·Updated 2 years ago· 814
📷Prompt_Styler
    • positive_prompt
    • negative_prompt
    promptsci-fi, closeup portrait photo of a man img wearing the sunglasses in Iron man suit, face, slim body, high quality, film grain
    negative_promptasymmetry, worst quality, low quality, illustration, 3d, 2d, painting, cartoons, sketch), open mouth
    style_namePhotographic (Default)

    Prompt_Styler is the pack's prompt-dressing node: you give it a prompt and it wraps that prompt in one of PhotoMaker's official style templates, then spits out a ready-to-use positive and negative pair. It's the closest thing this pack has to the A1111 "style" dropdown, and it arrived in the V2.5 rewrite specifically so the pack would play nice with other prompt tools like portrait-master and the general styler ecosystem.

    Mechanically it's dead simple: each style is a template string with a {prompt} placeholder plus a fixed negative. The node replaces the placeholder with your prompt and appends your own negative prompt to the style's baked-in one. The 11 styles are TencentARC's official set, so you're getting the same look presets the PhotoMaker demo used, not some community invention. Your inputs are the two prompts and the style dropdown:

    • prompt - your raw subject prompt, e.g. photo of a man img wearing sunglasses. Keep the img trigger word in here; the style templates won't add it for you.
    • negative_prompt - your own negatives, appended to whatever the style contributes.
    • style_name - the dropdown: (No style), Cinematic, Disney Charactor, Digital Art, Photographic (Default), Fantasy art, Neonpunk, Enhance, Comic book, Lowpoly, Line art.

    Outputs are positive_prompt and negative_prompt (both STRING), which wire straight into the positive and negative inputs of NEW_PhotoMaker_Generation. That's the natural loop: Styler → Generation. Since the generation node's prompt inputs are forceInput (you can't type into them), this node or any plain text node is the only way in.

    Style picks, with a take: Photographic is the safe default and what most people should start on - it's a light cinematic-photo wrapper that doesn't fight the identity. Disney Charactor is the crowd-pleaser if you want the Pixar look, and it pairs well with a Pixar LoRA through the pack's LoRALoader. Enhance and Line art are the ones I'd call overhyped for faces - they work, but they're generic. The style strength (which the separate generation node controls) is where the real tuning happens; this node is just the wrapper.

    Install is the standard pack setup - ComfyUI Manager (search "PhotoMaker") or:

    cd ComfyUI/custom_nodes
    git clone https://github.com/ZHO-ZHO-ZHO/ComfyUI-PhotoMaker-ZHO
    cd ComfyUI-PhotoMaker-ZHO
    pip install -r requirements.txt
    

    Restart ComfyUI and it appears in the 📷PhotoMaker category. It's a pure string node - no models, no VRAM, no failure modes beyond forgetting the img trigger word. And the standing caveat: this pack is an unmaintained early-2024 diffusers wrapper; the styler concept is still handy, but the native PhotoMaker nodes in ComfyUI core do the generation part lighter.

    Category📷PhotoMaker

    Inputs (3)

    NameTypeDefaultDescription
    promptSTRINGsci-fi, closeup portrait photo of a man img wearing the sunglasses in Iron man suit, face, slim body, high quality, film grain
    negative_promptSTRINGasymmetry, worst quality, low quality, illustration, 3d, 2d, painting, cartoons, sketch), open mouth
    style_nameCOMBOPhotographic (Default)11 options: (No style), Cinematic, Disney Charactor, Digital Art, Photographic (Default), Fantasy art, +5

    Outputs (2)

    NameTypeDescription
    positive_promptSTRING
    negative_promptSTRING