Local runtime
The engine that runs speech and voice workflows close to the work.
Product
Izwi brings the core voice stack into one product shape: runtime, API, CLI, desktop evaluation surface, and a path to supported deployment.
Product shape
The engine that runs speech and voice workflows close to the work.
Use /v1 endpoints if your team already integrates cloud AI APIs.
Pull models, start services, and operate the runtime locally.
A native surface for evaluation and prototyping.
Supported deployment path
Public downloads are for evaluation and self-serve use. Supported production deployment lives in the Enterprise Runtime path.
The core product is the runtime layer. The other surfaces help you evaluate it, integrate it, and operate it.
The engine that runs speech and voice workflows close to the work.
Use /v1 endpoints if your team already integrates cloud AI APIs.
Pull models, start services, and operate the runtime locally.
A native surface for evaluation and prototyping.
Supported deployment for controlled environments.
Izwi sits between raw model components and the workflow product built on top of them. It is the runtime layer, not the full application layer.
The clearer the product truth is, the easier the deployment story becomes.
Public downloads are for evaluation and self-serve use. Supported production deployment lives in the Enterprise Runtime path.
Start with the docs if you are integrating. Talk to us if you need a controlled deployment path.