TL;DR
Watch our video for step-by-step instructions to create your own ivrit.ai API endpoint.
More details
We created the ivrit Python package to help you user our models efficiently.
You can host them directly on your machine, or on a hosting service.
We typically use runpod, as it provides serverless inference coming in at ~0.03$ per transcribed hour.
If you want a hosted API, the easiest way is to use our latest image on the runpod hub, together with our Python package.
Need detailed instructions? watch our 7-minute video taking you through this with painstaking details.
Support
Found an issue? Have a feature request? Want to let us know it all sucks? We’re available at info@ivrit.ai.