Cozy Knit Sweater on Autumn Bench
FluxCozy Knit Sweater on Autumn Bench is a tested Flux Flux.1 dev prompt for social media work. It runs at Flux Flux.1 dev · Aspect ratio 4:5 · Steps 28 · CFG scale 3.5, and ships below with the exact prompt text, usage notes, and 2 pasteable variations.
Last tested: July 2026 with Flux Flux.1 dev
Part of the Flux prompt library · see more cozy prompts .
Prompt
cozy oversized knit sweater draped over a wooden park bench, scattered autumn leaves, a cup of coffee and a paperback book beside it, soft golden light, lifestyle flat-lay photography, warm hygge mood, 4:5
Settings
- Model
- Flux Flux.1 dev
- Aspect ratio
- 4:5
- Steps
- 28
- CFG scale
- 3.5
How to use this prompt
The sweater must look heavy and slightly rumpled; a neatly folded sweater reads retail. Draping one sleeve over the bench edge creates natural folds that catch the golden light. The coffee cup and book are lifestyle props that tell a story; remove them and it is just clothing. Keep leaves sparse and real; too many looks like a leaf pile. If the knit texture disappears, add "chunky cable knit" and lower cfg to 3.0. A neutral cream sweater reads more autumnal than bright red.
Variations to try
- Cottage version: place the bench against a white picket fence with ivy
- Moody version: foggy morning with muted colors and a wool scarf added
Explore collections
What to do next
For developers — copy this prompt as JSON
{
"title": "Cozy Knit Sweater on Autumn Bench",
"prompt": "cozy oversized knit sweater draped over a wooden park bench, scattered autumn leaves, a cup of coffee and a paperback book beside it, soft golden light, lifestyle flat-lay photography, warm hygge mood, 4:5",
"model": "Flux",
"model_version": "Flux.1 dev",
"settings": {
"aspect_ratio": "4:5",
"steps": 28,
"cfg": 3.5
},
"style_tags": [
"cozy",
"autumn"
],
"use_case_tags": [
"social-media",
"lifestyle-blog"
],
"source_url": "https://promptkoi.com/prompts/flux/cozy-knit-sweater-autumn/"
} The whole library is available as an open JSON dataset · developer docs.