Generate images and videos with ComfyUI 1 h

ComfyUI is a powerful, node-based interface for Stable Diffusion that transforms AI media generation into a visual programming experience.

Olares provides ComfyUI Shared to allow multiple users to share models, plugins, and workflow resources within the cluster. It also features ComfyUI Launcher, providing administrator users with a simple way to manage ComfyUI resources and runtime environments.

Learning objectives

• Install and configure the ComfyUI service.
• Use ComfyUI Launcher to download optimized model packages.
• Generate images using the Z-Image Turbo workflow.
• Generate videos using the Wan 2.2 model.

Prerequisites

Hardware

• Olares One connected to a stable network.
• Sufficient disk space to download models.

User permissions

• Admin privileges to install ComfyUI from the Market, and to start or stop the ComfyUI service for the cluster.

Step 1: Install ComfyUI

1. Open Market, and search for "ComfyUI".

2. Click Get, then Install, and wait for installation to complete.

   

After it finished, you can see two icons for ComfyUI on Launchpad:

• ComfyUI: The client interface for ComfyUI.
• ComfyUI Launcher: The core management dashboard. You must use this tool to start the ComfyUI service before you can use the client.

Step 2: Download model package

To generate content, you must first download the specific model package.

1. Open ComfyUI Launcher, and scroll down to Package installation.

2. For image generation, find Z-Image Turbo Package and click GET ALL to download necessary models.

3. For video generation, select Wan 2.2 Text to Video 14B Package and click GET ALL to download necessary models.

   

4. . When the download completes, click RESTART on the Home page for the changes to take effect.

Step 3: Start the ComfyUI service

1. In ComfyUI Launcher, click START in the upper-right corner.

   Initialization time

   The initial startup typically takes 10–20 seconds as the environment initializes.

2. Once the status changes to "Running," click OPEN. This will launch the ComfyUI client in a new browser tab.

Step 4: Generate images

This section uses the Z-Image Turbo workflow to help you get started.

1. In the ComfyUI client, click Templates in the left navigation bar to browse available templates.

2. Select Image as the generation type, and select the template: Z-Image-Turbo Text to Image.

3. When prompted that this workflow is missing models, simply close the window.

4. Prepend public/ to the filename in the Text to Image node. For example:

   • Default: ae.safetensors
   • Change to: public/ae.safetensors

   Shared model path

   ComfyUI in Olares uses a file structure that differs from the standard installation. This change allows models to be shared between ComfyUI and SD Web UI.

   

5. Update the text prompt in the same node to describe the image you want to generate.

6. Click Run in the toolbar to start generation.

Step 5: Generate videos

This section uses the Wan 2.2 workflow.

1. In ComfyUI, open the ComfyUI menu and select View > Browse Templates.

2. Under the GENERATION TYPE category, select Video.

3. Select the template: Wan 2.2 14B Text to Video.

4. Prepend public/ to the filename in each model loader node. For example:

   • Default: wan_2.1_vae.safetensors
   • Change to: public/wan_2.1_vae.safetensors

     Shared model path

     ComfyUI in Olares uses a file structure that differs from the standard installation. This change allows models to be shared between ComfyUI and SD Web UI.

   

5. Locate the CLIP Text Encode node and enter a detailed prompt for your video.

   A woman with long brown hair and light skin smiles at another woman with long blonde hair. The woman with brown hair wears a black jacket and has a small, barely noticeable mole on her right cheek. The camera angle is a close-up, focused on the woman with brown hair's face. The lighting is warm and natural, likely from the setting sun, casting a soft glow on the scene. The scene appears to be real-life footage.

6. Click Run in the toolbar to start generation. Video generation takes significantly longer than image generation.

Step 6: Download output files

You can download all output images and videos from Olares One to your local computer.

1. Open the Files app.
2. Navigate to the following directory:

   External/olares/ai/output/comfyui

   
3. Select the files you wish to save.
4. Right-click and select Download to save them to your local computer.

Resources

• ComfyUI official docs
• ComfyUI-nunchaku docs
• Manage ComfyUI using ComfyUI Launcher