I found it really easy to start using. The tutorial takes you through the basic features really quickly and gets you familiar with what Deepgram can do, the API and so on.
The response time of speech to text is a little high. Hindi support would be helpful too.
The best part is support for multiple languages. The transcripts are very accurate, and having it done as a cloud service helps us separate the concern of having accurate STT frmo the rest of our app.
1. Lack of documentation, or sometime the available document is incomplete or not up to date. 2. The default model does seem to pick the best model for the current audio and the user to pick a different model which performance better.
I found it really easy to start using. The tutorial takes you through the basic features really quickly and gets you familiar with what Deepgram can do, the API and so on.
The best part is support for multiple languages. The transcripts are very accurate, and having it done as a cloud service helps us separate the concern of having accurate STT frmo the rest of our app.
The response time of speech to text is a little high. Hindi support would be helpful too.
1. Lack of documentation, or sometime the available document is incomplete or not up to date. 2. The default model does seem to pick the best model for the current audio and the user to pick a different model which performance better.