Coqui STT

Deep-learning speech-to-text engine that runs offline on a wide range of devices.

Coqui STT is an open-source speech-to-text engine descended from Mozilla DeepSpeech, designed to run fast enough for real-time use on devices from datacenters down to a Raspberry Pi. It ships pretrained models in many languages plus tooling to train and fine-tune custom acoustic models for domain-specific vocabularies.