Luts Styler (Advanced)

Same misleading name as the basic LutsStyler, so let's be clear up front: this does not apply a LUT. No color lookup table, no .cube file, no image processing. The "LUTs" are ~101 named color-grade words - Cinematic, Sepia, Vibrant, Filmic, Muted - that get injected into your text prompt so the model generates something that already looks graded. The output is a STRING, not an IMAGE. If you wanted to remap the colors of a finished render, this is the wrong node entirely.

What "Advanced" adds is SDXL's dual-CLIP handling: the grade menu, with the positive prompt split across SDXL's two encoders and a control for where the negative lands.

How the "Advanced" plumbing works

SDXL has two text encoders, CLIP-G and CLIP-L. This node feeds them separately - text_positive_g to G, text_positive_l to L - and applies the chosen grade word to both, wrapped at the pack's strong default weight (~2.0). MileHighStyler is a fork of the classic SDXL Prompt Styler. negative_prompt_to (Both / G only / L only) routes the grade's built-in negative; Both by default.

The inputs and outputs that matter

• LUTs - the ~101-entry grade dropdown; none is off.
• text_positive_g / text_positive_l - your subject per encoder; same text in both if unsure.
• text_negative - your negative, with the grade's negative routed by negative_prompt_to.

log_prompt prints the assembled prompt so you can see what the grade added. Six outputs: text_positive_g, text_positive_l, merged text_positive, and text_negative_g / text_negative_l / merged text_negative - all STRING. Feed _g / _l into a CLIPTextEncodeSDXL, or take the merged text_positive / text_negative into a normal 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 an actual LUT. It changes your prompt, not your image's colors. No .cube, no color pipeline, no IMAGE output. For real grading, use a proper post-processing node.
• A prompt grade is a suggestion. "Cinematic" nudges; it won't clamp the palette like a real LUT. Results shift by checkpoint and seed.
• Advanced when basic would do. The G/L split only matters with a CLIPTextEncodeSDXL and different per-encoder text. Otherwise use the merged outputs, or the plain LutsStyler.
• Model era. Dual-CLIP is SDXL-only. Meaningless on the 2026 LLM-encoded models (Flux 2, Z-Image, Qwen); write "cinematic color grade" as plain words there.