Fat Mex Prompt Template

This node is a string template with a dropdown. You describe a scene - "woman in red bikini at the beach" - pick a style from eight photography presets, and it assembles a ready-to-use prompt. It exists because the pack targets AI influencer content, where you generate the same kind of image over and over and the prompt structure stops being interesting. If that's not your workflow, this node is skippable. If it is, the style presets are genuinely handy.

How it works

It's not a language model. There's no AI rewriting anything - it's eight hand-written templates:

• Instagram Photo - "instagram photo, {desc}, aesthetic, well-lit, high quality smartphone photo"
• Professional Portrait - studio lighting, bokeh background, 85mm lens
• Casual Selfie - "slightly imperfect, authentic feel"
• Plus Fashion Editorial, Fitness, Lifestyle Candid, Glamour, and Street Style.

Your description gets dropped into the {desc} slot, and if quality_boost is on (default true) it appends: , masterpiece, best quality, highly detailed, sharp focus, 8k uhd, professional photography.

Output is a single prompt STRING - wire it into any prompt input on the pack's samplers (or any text box in ComfyUI).

The honest take on quality_boost

Leave it on if you want, but know what it's doing: masterpiece, best quality, 8k uhd is SD 1.5/SDXL vocabulary. The KB's prompt-engineering doc has A/B tests showing these tags are inert on the LLM-encoded models this pack runs - Qwen and Flux encoders read prompts like instructions, and quality tags were never part of their training vocabulary. It's not harmful, it's just not doing the work it used to. The style part of the template is what actually shapes the output; the quality suffix is legacy seasoning.

So treat this node as a starting point: grab the style structure, then add your own specifics (camera, location, mood, subject details) rather than leaning on the tag soup. On a model like Klein or Qwen Image Edit, concrete description beats generic quality adverbs every time.

Install

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

Or ComfyUI Manager → search "Fat Mex Nodes" → Install → restart. No models, no downloads - pure text.

Where people get burned

• It doesn't write you a paragraph. The description field is your scene, and it's pasted verbatim. Feed it real detail ("woman in red bikini at the beach" is the default and it's thin); the template only adds style.
• Expecting AI creativity. If you want a model to invent the prompt for you, this isn't that. It's a formatter. Useful, boring, predictable - which is the point.