Small enough to fit on edge devices and existing servers without a hardware refresh
Small enough to fit on edge devices and existing servers without a hardware refresh
Transcribes as people speak, with sub-100 ms latency for partials
Transcribes as people speak, with sub-100 ms latency for partials
Trained on high entropy data to achieve industry best 3.10% WER
Trained on high entropy data to achieve industry best 3.10% WER