My 2-stage (<a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0">base</a> + <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0">refiner</a>) workflows for SDXL 1.0 in <a target="_blank" rel="ugc" href="https://github.com/comfyanonymous/ComfyUI">ComfyUI</a>, with separate prompts for text encoders. If you don't need LoRA support, separate seeds, CLIP controls, or hires fix - you can just grab basic v1.0 workflow. Works with bare ComfyUI (no custom nodes needed).SD+XL workflows are variants that can use previous generations SD models to provide initial scene concept.Links to SDXL 1.0 model files:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_base_1.0.safetensors">sd_xl_base_1.0.safetensors</a> or <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_base_1.0_0.9vae.safetensors">sd_xl_base_1.0_0.9vae.safetensors</a>* (6.94 GB),</li><li><a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/resolve/main/sd_xl_refiner_1.0.safetensors">sd_xl_refiner_1.0.safetensors</a> or <a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-refiner-1.0/resolve/main/sd_xl_refiner_1.0_0.9vae.safetensors">sd_xl_refiner_1.0_0.9vae.safetensors</a>* (6.08 GB),</li><li><a target="_blank" rel="ugc" href="https://huggingface.co/stabilityai/stable-diffusion-xl-base-1.0/resolve/main/sd_xl_offset_example-lora_1.0.safetensors">sd_xl_offset_example-lora_1.0.safetensors</a> (49.6 MB).</li></ul>* versions with fixed <a target="_blank" rel="ugc" href="https://github.com/huggingface/diffusers/issues/4310">issue #4310</a>.Link to ControlNet 1.1 model files (used in SD+XL v1.2):<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/lllyasviel/ControlNet-v1-1/tree/main">ControlNet 1.1</a>.</li></ul>Link to my SD 1.5 variant used in SD+XL workflow:<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/22511?modelVersionId=41974">MoonRide Mix 10</a> (you can replace it with any other SD variant you like).</li></ul>Parameters to play with:<ul><li>text in prompts,</li><li>width and height (but stick to combinations used during SDXL training - listed in notes section),</li><li>noise seeds,</li><li>steps ("end_at_step" is switching point between base and refiner - should be about 50% to 95% of total steps, depending on selected scheduler and samplers - look at preview image from base to check what is being passed to refiner stage),</li><li>scheduler,</li><li>sampler names,</li><li>cfg (lower - more creative, but also less defined),</li><li>CLIP Skip (-1 will treat prompts more literally, -3 will generalise),</li><li>LoRA strengths (0 means disabled).</li></ul>Cheers, and have fun!

MoonRide workflows for SDXL 1.0 in ComfyUI

Enhanced workflow using previous generation model (any variant of SD 1.x or 2.x will work) to provide initial concept of the scene for the SDXL models.

Files with SHA256: 0ab35b1b8d3be7b53c1c4537831f278cdc739045656087ea7e624aac914922f6