Welcome to this collection of workflows for ComfyUI, designed and optimized to work with various models and techniques. This repository is the ideal companion to the UmeAiRT ecosystem.
IMPORTANT: These workflows are designed to work with the specific ecosystem and custom nodes installed by the UmeAiRT ComfyUI Auto-Installer. If you are using a standard ComfyUI installation, you might encounter missing node errors.
Forget manual downloads. Use my custom loader node to keep your workflows automatically updated every time you start ComfyUI.
- Open ComfyUI Manager.
- Search for and install
ComfyUI-UmeAiRT-Sync. - Restart ComfyUI.
- Your workflows will appear automatically in
ComfyUI/user/default/workflows/UmeAiRT.
If you are setting up ComfyUI from scratch, the UmeAiRT installer automatically clones this entire repository into the correct folder during the installation process.
- Go to the
ComfyUI/user/default/workflowsdirectory. - Execute the command
git clone https://github.com/UmeAiRT/ComfyUI-Workflows.
To make finding files easier, workflows now follow a strict naming convention: Model_Function_Variant.json.
- Standard:
Flux_Txt2Img.json(Base version) - GGUF:
Flux_Txt2Img_GGUF.json(Optimized for low VRAM) - Nunchaku:
Flux_Txt2Img_Nunchaku.json(Specific backend)
Here is the list of available workflows, categorized by model family.
Workflows optimized for models in the FLUX family.
Flux_Txt2Img.json: Basic text-to-image generation.Flux_Img2Img.json: Transform an existing image.Flux_Kontext.json: Advanced contextual generation.
Flux_Inpaint.json: Modify or replace specific areas of an image.Flux_Outpaint.json: Extend an image beyond its original borders.
Flux_ControlNet_*.json: Precise control using Canny, Depth, HED, or OpenPose.Flux_LoRA.json: How to implement LoRA adapters.Flux_Upscale.json: High-fidelity upscaling.
Robust workflows for Stable Diffusion XL.
SDXL_Txt2Img.json: High-quality generation from prompt.SDXL_Img2Img.json: Image variation and restyling.
SDXL_Inpaint.json: Seamless object removal or addition.SDXL_Outpaint.json: Canvas expansion.
SDXL_ControlNet.json: Structure guidance.SDXL_LoraTester.json: Utility to quickly test LoRA weights.
Workflows for the Z-Image generation model.
Z-Image_Txt2Img.json: Standard generation.Z-Image_Img2Img.json: Image-to-image transformation.Z-Image_Inpaint.json: Native inpainting capabilities.
Workflows designed for the HiDream model.
HiDream_Txt2Img.json: Text-to-image generation.HiDream_Img2Img.json: Image transformation.HiDream_Inpaint.json: Native inpainting capabilities.
Workflows for the LTX-Video model.
LTXV_Txt2Video.json: Create short clips from text.LTXV_Img2Video.json: Animate static images.
Workflows for models in the WAN family.
Wan_Txt2Video.json: Standard text-to-video.Wan_Img2Video.json: Standard image-to-video.Wan_Face2Video.json: Character consistency in video.Wan_Video_Extension.json: Extend clip duration.Wan_Video2Loop.json: Create seamless loops.Wan_VACE.json: Advanced Video Composition/Editing.
Wan2.2_Txt2Video.json: High-quality 720p text-to-video generation.Wan2.2_Img2Video.json: Image animation with the new 2.2 model.Wan2.2_Txt2Img.json: Still image generation capabilities.Wan2.2_Img2Img.json: Image-to-image transformation.Wan2.2_StartEndFrame.json: Generate video bridging a start and end image.Wan2.2_FunControl.json: Advanced control features for video generation.
Utility tasks compatible with multiple models.
Img2Txt_Caption.json: Extract prompts from images (JoyTag/WD14).Video_Interpolation.json: Smooth out frame rates.Video_Upscaler.json: Simple upscaling for any video.Video_Merge.json: Stitch multiple video clips together.
Suggestions and contributions are welcome. If you have improvements for a workflow or have created a new one that could benefit the community, feel free to open an "Issue" or a "Pull Request".
This project is under the MIT License.