Version Detail
SDXL 1.0
🎨 Recommended txt2img Settings
Base Settings:
Sampler: DPM++ SDE
Scheduler: Normal / SGM Uniform (Pony) / Karras (XL/Photo)
Steps: 20–30
CFG: 3–10
Clip Skip: 2
Resolution: 832×1216 or 1024×1496
⚡ For faster results, you can use 2M SDE / 3M SDE, but note that quality will be slightly lower in Pony versions.
---
🌈 Rendering Quality Modes
Good Quality (Fast + Realistic):
DPM++ SDE + SGM Uniform
Steps: 8
CFG: 1.5–3 (Pony) / 1–2.5 (XL/Photo)
LoRA Strength: 1.0
Clip Skip: 2
Resolution: 832×1216 or 1024×1496
Best Quality (Natural Look):
LCM Exponential
Steps: 8
CFG: 1–1.5
LoRA Strength: 1.0
Clip Skip: 2
Resolution: 832×1216 or 1024×1496
---
🪄 Upscaling Recommendations
Option 1 – txt2img (Lightning 8-Step LoRA):
Enable Hires Fix
Upscaler: R-ESRGAN 4x+ recommended
Hires Steps: 3
Denoise: 0.4
Upscale: 1.5×
Hires CFG: 2
Clip Skip: 2
Option 2 – img2img (For Higher Quality & Speed):
Good Quality:
LoRA: Lightning 8-Step
DPM++ SDE + SGM Uniform
Upscale: 1.25–1.5× (XL) / 1.5–2× (Pony)
Steps: 3–12 depending on model type
CFG: 2
Denoise: 0.3–0.5
Clip Skip: 2
Best Quality:
LoRA: DMD2
LCM Exponential
Upscale: 1.25–1.5× (XL/Photo) / 1.5–2× (Pony)
Steps: 4–12 depending on version
CFG: 1
Denoise: 0.3–0.5
Clip Skip: 2
---
💡 Tips:
img2img gives multiple versions (thanks to seed variation).
Works great in A1111/Forge, and produces better upscaled detail than Hires Fix.
Ideal for portraits, fashion, and aesthetic renders.
---
⚠️ Important Disclaimer
> This model's training materials do not include any celebrity or unauthorized content.
Please do not use this model to post any content resembling celebrities.
Project Permissions
Use in TENSOR Online
As a online training base model on TENSOR
Use without crediting me
Share merges of this model
Use different permissions on merges
Use Permissions
Sell generated contents
Use on generation services
Sell this model or merges





