A personal project — not a company
Hi, I'm Edwin and I live in Sweden. I kept ending up with hours of voice memos, meeting recordings, and interviews that I wanted to read instead of re-listen to. The existing tools either felt clunky, charged a flat monthly fee for capacity I didn't need, or made the simple workflow of "record, transcribe, translate, save" surprisingly hard. So I built Transcribe for myself — pay-as-you-go, no subscription lock-in, accurate enough to actually trust.
After a few months of personal use I realised it was genuinely the fastest way I had to turn audio into text. Speaker diarisation, two translation modes, notebooks for organising — all the small features compounded. So I opened it up. Transcribe is still a one-person project that I run, maintain, and develop in evenings and weekends.
Record directly in the browser or upload an audio/video file (mp3, wav, m4a, mp4, mov, up to 500 MB). The audio is transcribed using a state-of-the-art speech-to-text model. You can translate the result, identify different speakers, and save the transcript as a note inside a notebook.
Every new account gets 1 free hour at signup. Need more? Buy an hour pack — they never expire. No subscriptions, no lock-in. Volume discounts apply on larger packs.
The site is operated and maintained by me personally. Hosting and transcription costs are covered by the hour-pack purchases. Because it's a personal project, there's no SLA — but I do my best to keep the site up and the transcriptions accurate. If something is broken, ping me via the contact form.
Found a bug, got feedback, or have a feature suggestion? Use the contact form →