Using in Desktop App
Get started with the WaveSpeed Desktop App to run models without writing code. It’s ideal for quick experiments, demos, and day-to-day content generation.
Prerequisites
- A WaveSpeedAI account and API key. See the Authentication guide or visit the access key page.
- macOS/Windows/Linux supported by the desktop build.
Install
- Open the GitHub repository for the desktop app:
https://github.com/WaveSpeedAI/wavespeed-desktop - Go to the Releases section and download the latest installer for your OS.
- Install and launch the app.
First-time setup
- Open the app.
- Paste your API key when prompted (you can find it at
https://wavespeed.ai/accesskey). - Save settings.
If you don’t have an API key yet, follow the Authentication guide.
Run your first model
- In the app, select a model (for example:
wavespeed-ai/z-image/turbo). - Enter your prompt (e.g., “Cat in space, cinematic lighting”).
- Click Run. The app will submit a job and show progress until it completes.
- When finished, you’ll see output URLs or previews you can download.
Tips:
- Start with short prompts and increase detail as needed.
- Use negative prompts or guidance fields (if available in the UI) to refine results.
Working with inputs
- Image-to-video or image-to-image: choose or drag a local file into the input field (the app will handle uploading behind the scenes).
- Video or audio inputs: similarly upload your source files, then configure any model-specific parameters.
Troubleshooting
- Invalid API key: re-check your key on
https://wavespeed.ai/accesskey. - Network/timeout: retry the job; consider trying again later if the service is busy.
- Model-specific errors: adjust inputs (prompt length, file size/format) per the model’s documentation in this site’s API pages.
Learn more
- Desktop App repository:
https://github.com/WaveSpeedAI/wavespeed-desktop - Full API reference and per-model docs: browse the items under “WaveSpeedAI API” and vendor sections in the sidebar.