1

00:00:00,000 --> 00:00:04,000

Transcribe without a third-party AI service

You can transcribe without a third-party AI service touching your recording, and here is the honest position: your file is uploaded to us, and it is never passed to OpenAI, Google, AWS or any other outside service. It is processed on hardware we own and physically operate, and the audio is deleted the moment your transcript is ready.

Choose a file

audio or video · your first file is free, no signup

encrypted · deleted right after transcription · no signup

What most transcription services actually do with your audio

A large share of transcription products are a website in front of somebody else's speech recognition. Your recording is uploaded, forwarded to a major cloud provider's transcription service, and the result is passed back to you. That means the audio has been handled by at least two companies, and their retention terms both apply.

This is rarely hidden, but it is rarely prominent either. It is worth reading any service's privacy policy for the specific question of whether your audio leaves their infrastructure, because that single fact matters more than most of the assurances that surround it.

Where this does and does not help you

What we can say precisely: no third-party service receives your recording, the machine doing the work is one we own and run, and the audio file is deleted as soon as the text exists. The transcript itself is kept so you can return to it, and we and our database provider can read that text, which we would rather state plainly than imply otherwise.

What this is not is local processing. If your requirement is that the audio never leaves your own machine at all, no hosted service of any kind meets it, including this one. In that situation run speech recognition yourself; our guide on doing that covers what it takes and where the difficulty lies.

Frequently asked questions

Is my audio used to train anything?

No. Your recording is transcribed and then deleted, and it is not used to train or improve any model. There is no third-party service in the path that could use it either.

How quickly is the audio actually deleted?

The moment your transcript is ready, normally within seconds of processing finishing and always within an hour. Only the transcript text remains, and you can delete that yourself at any time.

What if I need the audio never to leave my computer?

Then use no hosted service at all, ours included. Running speech recognition on your own machine is the only arrangement that meets that requirement, and our guide explains what it involves.