Transcribe Text

TranscribeText is a browser-based transcription service for turning audio and video into text without installing anything. Upload MP3, MP4, WAV, M4A, WEBM or MOV and it returns a timestamped transcript, exportable as plain TXT or as an SRT subtitle track.

Beyond the raw transcript it does speaker diarization — labelling who said what across a multi-person recording — and subtitle translation, so a recording can be captioned in a language other than the one it was spoken in. The vendor states support for more than 100 languages, accuracy above 90%, and roughly a few minutes of processing per hour of media on GPU-backed infrastructure, with AES-256 encryption on uploads.

The free tier covers 60 minutes and one file per batch. Paid plans lift both limits, and an enterprise tier adds API access and dedicated support.

It is aimed at the ordinary transcription jobs — lecture recordings, research interviews, meeting notes, podcast show notes and subtitling for video creators — rather than at a full editing suite. If you need to edit the audio itself, collaborate on a transcript with a team, or have transcription happen automatically during a call, a tool like Descript, Otter or Fireflies covers more ground. TranscribeText's narrower argument is that a timestamped SRT is a couple of clicks away from a file you already have.