About this item

Understands ไทย quickly using only the processor: about 0.2 – 0.6 seconds per sentence, no graphics card needed. Used automatically once it's your voice input language. Error rate in our tests: 10.3% ("Multilingual speech recognition": 14.3%). Model: Zipformer (GigaSpeech 2) License: Apache 2.0 (the full license downloads with the model).