Callout Relay model mirror
Byte-for-byte copies of the speech models the Callout Relay desktop app downloads, published as plain files so the app needs no archive decoder and can resume any file. Every file is pinned by SHA-256 in the app's catalogue and fetched from a fixed commit of this repo.
Nothing here was retrained or modified. Each model keeps the licence of its upstream release; see the source below and the sherpa-onnx documentation.
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support