Most meeting transcription products work by sending a bot into the call. That needs the organiser's permission, it announces itself to everyone, some companies block it outright, and it does nothing for a meeting that already happened or took place in a room. This page starts from the other end: you have a recording, and you want it as text.
It does not matter which tool made the file. Zoom, Microsoft Teams, Google Meet, Webex, Slack huddles, Discord and Skype all end up as MP4, M4A or WEBM, and a screen recorder gives the same. For the three most common platforms there are separate guides on where the recording is stored and how to download it, linked at the bottom of this page.
Board meetings, site meetings, parent-teacher conferences, interviews across a desk and housing association assemblies are recorded with a phone, and the voice memo it produces is a perfectly good source. Put the phone in the middle of the table rather than in front of one person, away from a laptop fan or a projector, and ask people not to shuffle papers next to it.
A room recording is harder for a speech model than a video call, because every voice reaches the microphone from a different distance and people interrupt each other more freely. Speakers close to the phone are transcribed cleanly; someone at the far end of a long table may need a second listen. The timestamps in the transcript take you straight to those places in the audio.
A transcript is the raw record, and minutes are what people actually read. With the text in front of you, writing them is a matter of searching rather than remembering: look for the words decided, agreed, deadline or the name of a project, and you land on the passage with the time it was said. Quotes can be copied exactly, which settles later arguments about who promised what.
Speaker labels are added to recordings you upload in a free account; the guest transcript on this page has timestamps only. The labels show where one voice ends and the next begins. They mark the voices as separate speakers rather than by name, so the first step in a free account is usually to check the names against the opening of the meeting, where people introduce themselves. On paid plans an automatic summary gives the main points of the meeting, and the AI chat answers questions such as what was decided about the budget.
Meetings contain salaries, client names, health matters and plans nobody has announced yet. The recording you upload is processed with speech models on servers we operate and is not forwarded to a third-party speech API. This applies equally to guest uploads and to accounts. It remains your responsibility to make sure the participants knew the meeting was being recorded.
As a guest you can read and copy a free preview of the transcript. A free account holds the full transcript, with an editor and a history of your meetings, and the transcripts you made as a guest move into it when you sign up. The free allowance is 5 hours across 10 recordings, shared between guest and account. SRT, VTT and ASS subtitle export, AI chat, summaries and export to PDF or Word are part of the paid plans.