Prompt Styler Subject

SDXLPromptStylerbySubject is one of the small utility nodes in wolfden's SDXL Prompt Styler pack. It keeps a short list of 11 subject categories - Animal, Art, Building, Child, Female, Food, Male, Nature, Technology, Vehicle. Pick the one that matches what you're making, type your subject, and the node wraps it in keywords tuned to that category.

It's a nudge, not a whole aesthetic. Where the big nodes throw thousands of art styles at you, this one just biases the prompt toward a subject type - useful as one link in a chain, or when you want a light, category-appropriate polish without committing to a specific movement.

How it works

Each category is a JSON entry with a {prompt} placeholder in its prompt field and a negative_prompt. Pick "Vehicle" and your text_positive slots into a template that emphasizes the things that make vehicle shots read well. No model, no API, no key - string substitution against a small lookup table. Because its output is just text, you can feed it into another styler node to layer setting or lighting on top.

The negative is baked into each preset and merges with anything you put in text_negative.

The inputs that matter

• text_positive - your subject, or another styler's output if you're chaining.
• style - the category, 11 options.
• text_negative - optional, merged with the preset's negative.

log_prompt (No/Yes) prints the assembled prompt to the console. Outputs are positive_prompt_text_g and negative_prompt_text_g, both strings; positive into the CLIP Text Encode feeding your positive conditioning, negative into the negative one. On SDXL the same string can go into both G and L of a CLIPTextEncodeSDXL.

Installing it

ComfyUI Manager: search SDXL Prompt Styler (wolfden), install, restart. Or manually:

cd ComfyUI/custom_nodes
git clone https://github.com/wolfden/ComfyUi_PromptStylers

Restart ComfyUI. It appears under the Style Prompts submenu. No models, no dependencies.

Where people get tripped up

Honestly, with only 11 broad categories this node does the least dramatic thing in the pack - the styling is subtle, and if you were expecting a big visual shift you'll be underwhelmed. That's by design; it's a modifier, best used alongside the setting, lighting, or depth nodes rather than on its own.

Chaining is where the small nodes earn their place, but each one you stack drags in its own inherited negative, and those add up into a generic block that can fight your positive on picky checkpoints. Flip log_prompt to Yes, read the final prompt, and trim. The presets are JSON in the pack folder, easy to open and edit.

Pack-wide caveat: these templates were built for SDXL and its CLIP encoder - great on SDXL, Illustrious, Pony, NoobAI. On 2026 LLM-encoded models the styling weakens and the inherited negative does nothing at CFG 1, where distilled checkpoints run, because the negative pass isn't computed. And if the node doesn't appear, restart fully and look under Style Prompts, not the root menu.