Model selection
Load and switch local models while testing speech workflows.
Izwi Desktop
Test private voice workflows on your own machine
Load supported models, test real audio, and compare speech workflows before writing application code. Processing stays on your machine.
Use desktop for
Evaluation before integration
Use Desktop to test model fit, workflow shape, and local behavior before writing integration code.
- test transcription models with real audio files
- compare speech settings before writing integration code
- explore local voice workflows without a hosted API
- prove that a workflow is worth moving into the runtime API
Audio capture
Use microphone or file input to inspect transcription behavior.
Workflow preview
Evaluate the shape of voice features before committing product code.
Ready to test the workflow locally?
Download the desktop app first, then move to the server when the workflow is ready for application integration.