Adjective Styler

AdjectiveStyler is exactly what it says: a dropdown of ~1,100 adjectives - abandoned, ablaze, abnormal, aberrant, all the way down the alphabet - that it welds onto your prompt at a strong weight. It's one of a small set of "grammatical" nodes in this pack (alongside the verb and preposition stylers), and I'll be straight with you: it's more curiosity than daily driver. Most people prompting seriously just type the adjective they want. The node's honest use is exploration - scroll a big list of descriptors, pick something you wouldn't have thought of, and see what an unexpected word does to a scene.

There's nothing wrong with that. Batch a few and you get cheap variety. Just don't expect a single injected adjective to transform an image the way an artist or camera cue can - an adjective is a light touch, and this node applies exactly one at a time.

How it works

MileHighStyler is a fork of the classic SDXL Prompt Styler. Each menu entry is a JSON template with a {prompt} slot; AdjectiveStyler drops your text_positive into the chosen adjective's template and wraps the style text at a strong default weight (~2.0, per the pack docs). Output is text, into your CLIP Text Encode - it rewrites your prompt string and nothing more.

The inputs and outputs that matter

• adjective - the ~1,100-entry dropdown. none is the off switch.
• text_positive - your subject; it lands in the template.
• text_negative - optional negative text.

log_prompt prints the finished prompt to the console so you can see the word (and its weight) in context. Outputs are text_positive and text_negative strings, into a CLIP Text Encode.

Installing it

ComfyUI Manager: search ComfyUI_MileHighStyler, install, restart. Or:

cd ComfyUI/custom_nodes && git clone https://github.com/TripleHeadedMonkey/ComfyUI_MileHighStyler

then restart. No models, no requirements.txt, no dependencies - Python and JSON. Nodes appear under ali1234/stylers.

Where people get burned

• Expecting too much from one word. An adjective at 2.0 is still just an adjective. It nudges; it doesn't restyle. If you want a real look, use the artist, camera, or big style menus instead.
• You could've just typed it. No shame in the menu, but this node's ceiling is "typing an adjective, from a list." Its value is discovery, not capability.
• Model era. The 2.0 weighting is CLIP-encoder syntax, discarded on the 2026 LLM-encoded models (Flux 2, Z-Image, Qwen). Use it on the SDXL family and SD1.5; on newer models, write the word plainly - an adjective works fine as literal text.
• Weird words, weird results. Some of these descriptors are abstract enough that the model has no visual handle for them and you'll get noise. That's the game; keep the seed fixed and A/B.