This is my Workflow to generate the Base-Image with Pony, Upscale it and refine it with a realistic SDXL Checkpoint. The Worklfow is think for user with good Hardware because it use 80 steps on each KSampler, but this is easy to change. With this Pony + SDXL-Mix you have the possibility to generate really realistic looking things like monster sex, that normally not possible with SDXL. With V1.1 the LoRAs add to the metadata for autodetection of civitai You can also use the same checkpoint for both and use it as normal Upscaling-Workflow. 🔧 Required Custom Nodes (better use ComfyUI-Manager to install)<ul><li>MultiLora Loader &amp; Lora Text Extractor <a target="_blank" rel="ugc" href="https://github.com/skfoo/ComfyUI-Coziness">https://github.com/skfoo/ComfyUI-Coziness</a></li><li>ComfyUI-Image-Saver (alexopus fork) <a target="_blank" rel="ugc" href="https://github.com/alexopus/ComfyUI-Image-Saver">https://github.com/alexopus/ComfyUI-Image-Saver</a></li><li>Impact Pack (ImpactWildcardEncode) <a target="_blank" rel="ugc" href="https://github.com/ltdrdata/ComfyUI-Impact-Pack">https://github.com/ltdrdata/ComfyUI-Impact-Pack</a></li><li>Efficiency Nodes <a target="_blank" rel="ugc" href="https://github.com/jags111/efficiency-nodes-comfyui">https://github.com/jags111/efficiency-nodes-comfyui</a></li><li>Ultimate SD Upscale / Latent Upscaler <a target="_blank" rel="ugc" href="https://github.com/City96/SD-Latent-Upscaler">https://github.com/City96/SD-Latent-Upscaler</a> or <a target="_blank" rel="ugc" href="https://github.com/ttl/ComfyUI_NNLatentUpscale">https://github.com/ttl/ComfyUI_NNLatentUpscale</a></li><li>CryTools (Crystools) <a target="_blank" rel="ugc" href="https://github.com/crystian/ComfyUI-Crystools">https://github.com/crystian/ComfyUI-Crystools</a></li></ul>

txt2img Multi Checkpoint + Multi LoRA + 2K-8K Upscaler + Aspect Selector

replace missing nodes Scaler Select 2K-8K Save Image with LoRAs

Files with SHA256: ad708180d97e613c3a88939b7713517447c44ecb9bbff828f01f7f9a7120b0ec