About this version:

🔧 v2LS (LORA Safe) – Recommended Settings

🧠 VAE

• Baked into the model

🌀 Sampler

• Recommended: DPM++ 3M SDE or Euler a

📅 Scheduler

• Recommended: SGM Uniform or Simple

⚙️ Steps

• Best results around: 30–40

📏 CFG Scale

• Suggested: 4-8

🧰 Extras

• ✅ Use Face Detailer and Highres Fix as needed

• 🆙 Recommended Upscaler: DAT-4x

• 🔄 Denoise Strength: 0.4 or lower for both Face Detailer and Highres Fix

💬 Other settings can work too — these are just my personal go-tos for balancing quality and generation time.

📂 Want real examples?
All sample images include the exact prompts and full ComfyUI workflows used – feel free to try them out or build from them!

📸 Model Info

• What’s Special?
  LS stands for LORA Safe —this version is pure SDXL for maximum compatibility with LORAs.

  • 🛡️ Use this if you notice artifacts or image quality issues in v1 or v2 with LORAs

  • 🔄 Great for users who rely on LORA add-ons

• Merge:

  • realvisxlV50_v50Bakedvae

  • Juggernaut XL Ragnarok

❤️ Stay Connected & Support

Catch early previews or support model development on Ko-Fi:
👉 ko-fi.com/almascar ☕✨
Every bit of support helps me keep creating cool new models — thank you!

this is not my model, i take no credit. All credit goes to creator @ https://civitai.com/models/1761696?modelVersionId=2070007