VOICE / SPEECH / MUSIC
Deepgram
Paid links. We earn a commission if you buy through links on this page. It never changes our ranking — see how we rank.
Pricing
We do not reprint prices. They change too often for that to be honest, and a number on our page with no date is worse than sending you to the source.
How we rated it
What it is
Speech-to-text through an API, built for speed: it transcribes recorded audio quickly and handles live streams with low enough latency to use in a phone call or a meeting.
Who it is for
Developers putting transcription inside a product — a call centre, a meeting tool, a voice interface — rather than people who want to transcribe a file now and then.
Where it is strongest
Live transcription where the delay is the whole problem. If you need words appearing while someone is still speaking, this is the category it leads.
Where it falls short
No interface worth the name; it is an API and nothing else. Accuracy on heavy accents and crosstalk trails the best offline models. And you are billed by the minute, which makes a runaway process expensive quietly.
Our verdict
Recommended when transcription is a feature of something you are building. If you just want a meeting written up, Otter or Fathom do the whole job and this does one part of it.
Editorial assessment. We earn a commission on some links; it does not change the ranking.