Getting started: your first transcript

Upload, or paste a link

Drop an audio or video file onto the upload panel — MP3, WAV, M4A, MP4, MOV and a dozen other formats all work. You do not need an account to start: upload a file and the transcript appears.

If the recording lives online, paste its address instead. WhoScribe fetches the audio and queues it like any other file.

Three settings worth a moment

Language. Leave it on automatic and the model detects the language itself. Setting it explicitly helps in two cases: very short recordings, and audio where two languages are mixed and you want one of them.

Speaker labels. Turn this on for interviews, meetings and panels. The transcript then arrives split by speaker rather than as one block of text. You can rename a speaker once and every line follows.

Quality. Balanced is the default and the right answer most of the time. Fast returns sooner on casual recordings; Best is worth the extra wait on difficult audio.

While it runs, and after

Transcription happens in the background — close the tab if you like, the file keeps processing and waits for you in your workspace.

When it is done you can read along with the audio, fix a word inline, rename speakers, and translate the whole transcript into another language.

Then export: DOCX for editing, PDF for sharing, TXT for plain text, SRT or VTT for subtitles. Or send a read-only link to someone who has no account.