Record the meeting so it can be transcribed
Tell everyone at the start that you are recording, and follow your organization's rules and the consent rules where the participants are. Then give the recording a fair chance:
- One recorder, in the middle. A phone flat on the table between speakers picks up a small room. In a large room, put it near the people who will talk most.
- Keep the table quiet. Typing, paper and cups next to the microphone are louder in a recording than they seem in the room.
- One conversation at a time. When people talk over each other, the transcript usually keeps one voice, so crosstalk is where words go missing.
- Say names when they matter. "Over to you, Priya" makes the transcript easier to label and check later.
For a video call, record in the meeting app and keep the file. Some apps transcribe recordings themselves: Microsoft Teams starts a transcript when you record and lets organizers download it as a Word or VTT file, and Zoom transcribes cloud recordings on its paid plans. Start with that transcript if you have one. Use the recording here when no transcript was made, when you need to correct it against the audio, or when you want other formats.
Transcribe and label speakers
Open the transcription studio, choose the recording and pick the spoken language. Audio files work, and so do videos such as MP4 and MOV call recordings: StemVocal transcribes their sound.
To see who said what, choose how many people speak before you start, up to 8. Speaker labels use credits. They are estimates, most reliable when people take clear turns, so check short replies and interruptions. Rename each label once the transcript is ready, and the name follows every turn.
A free job covers a meeting up to 30 minutes. Longer meetings, up to 3 hours, use one credit for each started 30 minutes, so a 90-minute meeting uses three. Weekly meetings can go in together: choose several files to make a batch of up to 20, then download every transcript in one ZIP.
From transcript to minutes
A transcript is the record; minutes are the summary people act on. Work through the transcript once with these in mind:
- Search for decision words such as "agree", "decide", "approve" and "next week", and for each person's name. Click a result to hear the moment and confirm it.
- Correct the names and terms that matter. Find & replace fixes a misheard project name everywhere at once, and one Undo restores it.
- Download Word with timestamps. Each turn starts with its time and speaker, so every line of the minutes can point back to the moment it came from.
- Write the minutes under short headings: date and attendees, decisions, actions with an owner and a due date, and open questions.
| You need | Download |
|---|---|
| A document to turn into minutes | Word, with timestamps |
| Notes to paste into an email or wiki | Plain text, or Copy text |
| A list of turns to filter by speaker | CSV, with start and end seconds |
| Captions for the recorded call | SRT or VTT |
| A copy to file or print | Print / PDF |
Attendees who work in another language can get a translation. A paid transcript can be translated from English into 15 languages, or from any of them into English. Review machine translation before you share it.
Keep meeting recordings private
Meetings often include confidential details. When a recording is uploaded for transcription, it goes to private storage with access for 24 hours and deletion normally within another 24 hours; you can delete it sooner from the editor. A free transcription that runs in your browser keeps the recording on your device. Edited transcripts save in this browser under Saved transcripts until you delete them, so download the minutes you need to keep. For one-on-one conversations, see interview transcription.
Sources: Microsoft Support, "Start, stop, and download live transcripts in Microsoft Teams meetings"; Zoom Support, "Using audio transcription for cloud recordings". Checked October 11, 2026.
Good to know
Can it transcribe a Zoom, Teams or Meet recording?
Yes, as a file. Save or download the recording, an MP4 video or an M4A audio file for example, and choose it in the studio. The transcript is made from its sound.
How many speakers can it label?
Up to 8. Speaker labels use credits; choose the most voices you expect before you start, then rename each label. Labels are estimates and work best when people take clear turns.
Can I transcribe several meetings at once?
Yes. Choose up to 20 files to make a batch. Free files run one at a time, each using one of your free jobs; with credits, files run two at a time. Download all makes one ZIP.
Is it free?
Yes, for recordings up to 30 minutes. You get 2 free StemVocal jobs each UTC day without an account, shared with the other StemVocal tools. Longer meetings, up to 3 hours, use one credit per started 30 minutes.