Using Image Generation Models
GPUStack supports deploying and running Image Generation Models. This guide walks you through deploying the Z-Image-Turbo text-to-image model and generating images from textual prompts in the GPUStack UI.
Prerequisites
Before you begin, ensure that you have the following:
- A GPU with at least 24 GB of VRAM, e.g., NVIDIA RTX 4090.
- GPUStack installed and running. If not, refer to the Quickstart Guide.
- Access to Hugging Face or ModelScope to download the model files.
Step 1: Deploy the Z-Image-Turbo Model
Follow these steps to deploy the model:
- Navigate to the
Model Catalogpage in the GPUStack UI. - Select
Imagein the category filter. - Click the
Z-Image-Turbomodel card to open the deployment page. - Keep the default settings and click the
Savebutton to deploy.
After deployment, you can monitor the status on the Deployments page. Once the deployment is successful, click the ellipsis icon of the deployment and select Open in Playground to start using the model in the Playground.
Step 2: Use the Model for Image Generation
In the Image playground:
- Select the
Sizedropdown and choose the desired image resolution. - Enter a prompt describing the image you want to generate. You can click the
Random Promptbutton to get a random prompt for inspiration. - Click the
Submitbutton to create the image.
The generated image will be displayed in the UI. Results vary with randomness, seeds, and prompt details.
Conclusion
With this setup, you can generate unique and visually compelling images from textual prompts using the Z-Image-Turbo model. Experiment with different prompts and settings to create a wide variety of images!


