Related Posts
Latest Posts ⚡
Jetson Generative AI – Stable Diffusion WebUI
The world of AI art generation has been revolutionized by diffusion models—Stable Diffusion WebUI brings the power of text-to-image generation directly to your Jetson device. Generate stunning, high-quality images from simple text prompts using AUTOMATIC1111’s popular web interface in real time on your Jetson hardware.
In this article you’ll learn how to run Stable Diffusion WebUI on Jetson using optimized Docker containers and NVIDIA’s acceleration libraries.
Requirements
|
Hardware / Software
|
Notes
|
|---|---|
|
Jetson AGX Orin
|
Orin AGX 64GB recommended for best performance
|
|
JetPack 5 (L4T r35.x) or JetPack 6 (L4T r36.x)
|
Needed for latest pre-built containers
|
|
NVMe SSD storage
|
Highly recommended for speed and space
|
Storage Requirements
To run Stable Diffusion WebUI effectively you’ll need adequate storage space:
– 6.8 GB for container image
– 4.1 GB</strong> for SD 1.5 model
– 6.8 GB for SDXL model (if using Stable Diffusion XL)
– Additional space for generated images and custom models
Tip: An NVMe SSD is highly recommended for faster model loading and image generation times.
Step-by-Step Setup
1. Clone the repository
Copy to Clipboard
2. Install jetson-containers
Copy to Clipboard
3. Update system packages
Copy to Clipboard
4. Install helper Python packages
Copy to Clipboard
5. Verify GPU availability
Copy to Clipboard
Ensure your Jetson GPU is detected and CUDA drivers are properly installed.
6. Start Stable Diffusion WebUI
Copy to Clipboard
The container will automatically start the webserver with optimized settings:
Copy to Clipboard
The first run downloads the model checkpoint (≈ 4.1 GB for SD 1.5).
7. Open the Web UI
When the console prints:
Running on local URL: http://0.0.0.0:7860
On-device: open a browser on the Jetson and visit `http://localhost:7860`
Remote: replace `<jetson-ip>` with your board’s address:
http://<jetson-ip>:7860
8. Generate your first image
Enter a text prompt in the main text box, such as:
– “A serene mountain landscape at sunset”
– “A futuristic city with flying cars”
– “A cute robot reading a book”
Click Generate and watch as your Jetson creates art from your words!
Stable Diffusion in Action
Below are demonstration videos showing Stable Diffusion’s capabilities for text-to-image generation on Jetson devices.
Text-to-Image Generation
Experience the magic of AI art creation with Stable Diffusion WebUI. Watch as simple text prompts transform into detailed, creative images in real-time on your Jetson device.
Troubleshooting
|
Issue
|
Fix
|
|---|---|
|
Out of memory errors
|
Use lower resolution (512×512) or enable low VRAM mode
|
|
Slow generation times
|
Ensure NVMe SSD is used; consider using smaller models
|
|
Container fails to start
|
Check JetPack version compatibility and available disk space
|
|
WebUI not accessible
|
Verify firewall settings and that port 7860 is not blocked
|
For more information about Stable Diffusion optimizations and advanced configurations, visit the jetson-containers GitHub repository and explore the Stable Diffusion documentation.