Launched this week
JOBE Vox
Affordable transcription API with speaker diarization
4 followers
Affordable transcription API with speaker diarization
4 followers
Speech-to-text API with speaker labels and [hh:mm:ss] timestamps out of the box, for $0.08 per audio hour. Send a public audio URL, get a diarized transcript.






I built an audio transcription service to solve a problem I was facing with another product of mine, StockCast.
I needed to transcribe earnings calls — long recordings with multiple speakers, noisy conference lines, and sometimes pretty bad audio quality.
Existing transcription services either struggled with these recordings or were too expensive for my use case.
So I built my own.
After using it successfully in production for the past 3 months, I decided to turn it into a standalone API for other developers. That's how JOBE Vox was born.
What it offers:
Speaker diarization (who said what)
Timestamps for each speaker turn
Support for long recordings and challenging audio
Simple API — just provide a public audio URL
$0.08 per audio hour, with no subscription
A note about pricing: I'm bootstrapping this project entirely on my own, so I can't offer a free trial right now. Instead, I made the minimum credit purchase just $0.50 (enough for over 6 hours of audio). Most of that first payment goes to Stripe fees anyway.
I'd love to get feedback from developers using it with different languages and audio sources.
If you give it a try, please let me know:
What language did you transcribe?
How was the transcription and speaker identification quality?
What's your use case?
You can try it here: https://vox.jobe.works
If you find a recording where Vox struggles, I'd especially love to hear about it. Those are the cases that help me improve the service.
Thanks!