1. Workflow Overview

This workflow enables one-click multi-style image generation with:

• 80s book illustration

• Pixel art

• B&W line drawing

• Master photography

Key Features:

• Batch prompt management via CR Prompt List

• Auto-generation of 4 styles using forLoop

• Dual-model system (AWPainting + Flux)

2. Components Breakdown

Core Nodes:

1. Style Switcher

   • easy anythingIndexSwitch: Style prompt selector

   • Text String: Preset style descriptions

2. Dual-Model Architecture

   • AWPainting_v1.3: Main artistic model

   • Flux.1: Detail enhancement

3. Advanced Sampling

   • SamplerCustomAdvanced: Custom noise/guidance

   • BasicScheduler: Optimized 20-step sampling

Dependencies:

• Required plugins:

  git clone https://github.com/rgthree/rgthree-comfy
  git clone https://github.com/ltdrdata/ComfyUI-Impact-Pack

3. Workflow Structure

Processing Paths:

1. Basic Path (Quick):
   AWPainting → KSampler → Direct output

2. Advanced Path (Precise):
   Flux → FluxGuidance → ConditioningConcat → Multi-style output

Loop Logic:（Python）

for i in range(4):  # Iterate 4 styles
    style = switch_style(i)
    generate_image(style)

4. Inputs & Outputs

Input Parameters:

• Required: Base prompt (e.g. "beautiful scenery")

• Optional:

  • Resolution (default 1024x1024)

  • Style weight (via FluxGuidance)

Outputs:

• 4 styled images (PNG)

• Auto-saved to ComfyUI/output

• Metadata records style parameters

5. Notes

• Hardware:

  • VRAM: ≥8GB (for 1024x1024)

  • Recommended: Run with --highvram

• Troubleshooting:（Python）

  # Error: CLIP text encode failed
  Fix: Check for special chars (like <>) in prompts
  
  # Error: VRAM overload
  Fix: Reduce resolution or use `--medvram`

• Extensions:

  1. Add LoraLoader for style LoRAs

  2. Use ImageUpscaleWithModel for higher resolution

Pro Tip: For anime styles, replace AWPainting with RevAnimated and set CFG=7.0.