Audiomenti.
How it works Languages API Pricing Log in

Privacy

Effective 2026-11-17. This is what Audiomenti does with your recordings, your transcripts and your account, written to be read.

1. The short version2. Who is responsible3. Your audio, step by step4. Everything else we hold5. Legal basis6. How long we keep it7. Who processes it for us8. Transfers outside the EU9. Cookies and your browser10. Your rights, and how to delete things11. Other people in your recordings12. Security13. Children14. Changes

1. The short version

  • Only the sound is used. From a video we transcribe the audio track; the picture is never looked at.
  • Audio is kept only while it is being transcribed. The moment a file has become text, the recording is deleted from our storage. A file that fails is deleted too.
  • The transcription runs on a rented GPU. We send the audio to Replicate, a third-party provider in the United States, whose machine runs the speech model. Replicate removes what it received, and the result, within an hour.
  • The text stays with you. Transcripts are kept on your account so you can come back for them, in any format, until you delete them.
  • Delete means wiped. Delete a transcript, or your whole account, and it is removed from our database and storage for good. There is no bin and no archive copy.
  • We do not train anything on your recordings or transcripts, we do not read them, and we do not sell or share them.

2. Who is responsible

Audiomenti is run by Eggvelop ApS, CVR 44287463, Denmark, which is the data controller for your account data. For what is said inside your recordings, you decide why and how they are transcribed, and we process them on your instructions: you are the controller and we are your processor. Reach us at contact@audiomenti.com.

3. Your audio, step by step

  1. In your browser. When you drop files, a folder or a zip, your browser reads them on your own device to count them and measure how long they are. Nothing has been sent to us at this point. That is how you get an exact price before anything leaves your computer.
  2. Upload. When you press the button to go on, the audio files are uploaded over https to our storage at Cloudflare, together with each file's name, size and measured length. From a zip, only the audio files inside are uploaded; documents, pictures and anything else in it never leave your device.
  3. Before payment. On the site nothing is uploaded until you have paid. Through the API you can upload first; an upload that is not paid for within 24 hours is deleted automatically, files and all.
  4. Transcription. Once paid, each file is handed to Replicate through a private, unguessable link that only works while that file is being transcribed. Replicate's GPU downloads the audio, runs the speech model and returns the text with timestamps. Replicate receives the audio and nothing else: not your email, not your file names, not your account.
  5. Deletion of the audio. As soon as the text for a file has arrived, that file's audio is deleted from our storage and its link stops working. Replicate deletes its copy of the input and the output within an hour, which is its default for work sent through its API.
  6. The text. The transcript is stored on your account, with the original file name, so you can read it and download it again in any format.

4. Everything else we hold

whatwhy
Email addressTo log you in with a one-time link, to tell you a transcript is ready (you can switch that off on your account page), and to send receipts and notices about the service.
API keyThe key is your account. It is stored in your browser when you are logged in and sent with each request.
Hours and ledgerYour balance and every change to it: purchases, jobs, refunds.
Jobs and filesFor each job: language, output format, when it was made and finished. For each file: its name and folder path, size, length, word count and status.
TranscriptsThe text and its timestamps, until you delete them.
PurchasesWhat was bought, for how much and when, and the Stripe customer and subscription ids. Card details never reach us.
Request countsHow many API requests your key made per day, to apply your plan's limit.
Terms acceptanceThe date you accepted the terms, and which version.
Emails you send usMail to any audiomenti.com address (and our replies) is kept in our support inbox, with its attachments, so we can answer and follow up. When you write, we see the account that belongs to your address, to answer questions about it.

We do not keep your IP address in our own records. Cloudflare, which serves the site, processes it to deliver pages and to stop abuse.

5. Legal basis

  • Performing the contract (GDPR article 6(1)(b)): the email, key, ledger, jobs, files, transcripts and Stripe ids are needed to provide what you paid for.
  • Consent (article 6(1)(a)): analytics in your browser, only after you accept them in the cookie dialog. Withdraw at any time under "Cookie settings" in the footer.
  • Legitimate interest (article 6(1)(f)): rate limiting and abuse prevention, so the service stays up for everyone.
  • Legal obligation (article 6(1)(c)): payment records are kept for the period the Danish Bookkeeping Act requires.

6. How long we keep it

  • Audio: from upload until that file has been transcribed, normally minutes. Unpaid uploads are deleted after 24 hours.
  • Audio at Replicate: removed within an hour of the transcription.
  • Transcripts, file names and job details: until you delete the job or your account.
  • Account data (email, key, plan, ledger): for as long as the account exists.
  • Login links expire after 30 minutes and are pruned afterwards.
  • Daily request counts: 90 days.
  • Emails you send us, and our replies: 400 days after the last message.
  • Analytics, if you accepted them: up to 12 months, on our own server.
  • Cloudflare's technical logs are held by Cloudflare for a short, rolling period, currently up to seven days.
  • Payment records at Stripe and in our bookkeeping: five years after the end of the financial year, as Danish law requires, even if the account is deleted.

7. Who processes it for us

These companies handle data on our behalf, and only to provide their part of the service:

processorwhat forwhere
Cloudflare, Inc.Hosts the site and the API. Stores accounts, transcripts and, while a file is being transcribed, its audio. Sends our emails (login links, receipts, transcript-ready notices) and receives the emails you send us.Global network; EU data centres serve EU visitors. US company.
Replicate, LLCThe rented GPU that runs the speech model. It receives the audio of each file being transcribed and returns the text. It receives nothing that identifies you.United States.
Stripe Payments Europe, Ltd.Card payments, subscriptions, invoices and the billing portal. Stripe is an independent controller for the payment itself.Ireland, with processing in the US.
Hetzner Online GmbHThe server that runs Umami, open-source analytics software we operate ourselves, for the analytics you accepted. No analytics company is involved.Finland.

We do not sell, rent or share personal data with anyone else, and we do not use advertising services. The speech model is an open model (NVIDIA Parakeet) that we run ourselves on Replicate's hardware; your audio is not sent to the model's maker.

8. Transfers outside the EU

Cloudflare, Replicate and Stripe are US companies or process data in the US. Those transfers rely on the EU-US Data Privacy Framework where the company is certified, and otherwise on the European Commission's standard contractual clauses. What crosses the border: your email and account records at Cloudflare and Stripe, and at Replicate the audio you asked us to transcribe, for the time the transcription takes and up to an hour after. If your recordings must not leave the EU, Audiomenti is not the right tool for them, and we would rather say so here than have you find out later.

9. Cookies and your browser

Audiomenti sets no advertising cookies. Your browser stores your API key when you are logged in, and the ids of transcripts you started without an account, so the site can show them to you again. Analytics run only if you accept them. The full list, with what each item is for and how long it lasts, is on the cookies page. Stripe's checkout and billing pages set their own cookies under Stripe's policy.

10. Your rights, and how to delete things

Most of it you can do yourself, right now:

  • Delete a transcript: open it and press Delete, or call DELETE /api/v1/jobs/:id. The text, the file names and any audio still waiting are removed from our database and storage at once and for good.
  • Delete your account: on your account page. Every transcript, the ledger, the key and your email are wiped. A running subscription is cancelled. Payment records stay at Stripe for the period the law requires.
  • Take your data with you: every transcript downloads in open formats (text, SRT, VTT, JSON, Markdown), singly or as one zip.

You also have the right to see what we hold about you, to have it corrected, to restrict or object to its processing, and to complain to a supervisory authority, in Denmark Datatilsynet. Write to contact@audiomenti.com and we answer within 30 days, usually much sooner.

11. Other people in your recordings

Recordings usually contain other people's voices and what they said. You are responsible for having the right to transcribe them: a lawful reason under the rules that apply to you, and consent where the law asks for it. We process such recordings only to produce the transcript you ordered. If you are a company that needs a written data processing agreement for this, write to us.

12. Security

Everything travels over https. Each audio file sits behind its own random link that is given only to the transcription machine and dies when the file is done. Keys are random and login is by one-time link, so there is no password to leak. Account data lives in Cloudflare's managed database with access limited to the service itself. Card details never touch our systems. If a breach ever affected your data we would tell you, and the authority, as the law requires.

13. Children

Audiomenti is not directed at children and accounts require the holder to be at least 18. If you believe a child has created an account, write to us and we will delete it.

14. Changes

The date at the top is the effective date of the current version. For changes that affect what we collect or why, we email every account before they apply.

See also the terms of service. Contact: contact@audiomenti.com.

Audiomenti.

Your recordings as text. One file or thousands of hours, one text file back for each.

Product Transcribe audio How it works Languages Pricing API
Audio to text Turn audio into textBulk transcriptionOne-time transcriptionMultiple files at onceTranscribe a zipMP3 to textAudio to SRT subtitlesTranscription API
Your data Privacy Terms Cookies
Network podmenti.com bookmenti.com eggvelop.com Contact contact@audiomenti.com

© 2026 Eggvelop ApS, CVR 44287463. Audiomenti is part of the Eggvelop network.