How to transcribe a Microsoft Teams meeting with speaker labels

By Alex Topilski, Founder

2026-08-03

Every Microsoft Teams meeting you record is saved as an .mp4 file in OneDrive or SharePoint, usually within a few minutes of the meeting ending. Downloading that file takes about 30 seconds. Turning it into a transcript with a speaker label on every line - so you can tell who said what without listening back - takes roughly the same time as the meeting itself ran. This guide covers exactly how to do both, when Teams' own transcript feature is worth using instead, and what to do if your organisation has not enabled recording at all.

How Teams stores recordings

Since 2021, Teams has saved meeting recordings to Microsoft 365 cloud storage rather than a local folder. Where the file lands depends on the meeting type:

The file format is always .mp4. A 1-hour meeting at Teams' default quality is typically 400-700 MB - large enough that the audio-only route (below) often uploads noticeably faster. Your IT admin may set a retention policy that deletes recordings after a fixed number of days; if your organisation enforces one, download the file before that window closes.

If you cannot see the recording - no thumbnail in Chat, no file in OneDrive - the most common reasons are: the meeting ended before recording completed, recording was not enabled by your IT admin, or your licence tier does not include it. Meeting recording requires Microsoft 365 Business Basic or above; the free Teams plan does not allow recording.

Step 1. Download the recording from OneDrive or SharePoint

There are two fast routes depending on where you want to find the file:

• From Teams Chat. Open the chat for the meeting, find the recording thumbnail (it appears alongside the meeting details), click it to open the OneDrive or SharePoint preview, then select Download in the top bar. The file saves as .mp4.
• From OneDrive directly. Go to onedrive.live.com or your organisation's OneDrive, open the Recordings folder, right-click the file and choose Download. Channel meeting recordings live in the channel's SharePoint site: Teams → the channel → Files tab → Recordings.

If only the audio is needed and you want a smaller file, any free audio editor - Audacity is a common choice - can export just the audio track as an .m4a or .mp3 in under a minute. An hour of audio is typically 20-60 MB, versus 400-700 MB for the full video, and uploads proportionally faster. Either format works for transcription; the full .mp4 is fine as-is.

Step 2. Upload to WhoSayAh

Drop the .mp4 (or extracted audio) on the WhoSayAh home page. The upload dialog shows the file's name, size and detected length. No account is needed; files up to 100 MB and 60 minutes are transcribed free, and both limits are checked at upload time, so an over-limit file is rejected within seconds before any waiting begins.

If your Teams recording is longer than 60 minutes, split it first. Most free audio editors can cut a file by time marker without re-encoding; Audacity's Export Audio command with a selection range is the quickest route. You get two files to upload separately, each producing its own transcript, which you combine after.

The dialog asks two questions before you can press Transcribe:

Language. Leave Detect automatically on unless the meeting was very short (under 2 minutes of actual speech) or the audio is particularly noisy. For a multilingual meeting, pick the language that dominated. Auto-detection reads the audio signal, so it has less to work with from a bad recording than from a clean one.

Speaker count. Keep Label speakers on and enter how many people actually spoke, not how many joined. A 20-person all-hands where 4 people spoke is a 4-speaker recording. If you are genuinely unsure, I don't know - detect it gives a reasonable result, but the exact count is the single setting that most improves attribution accuracy. Then press Transcribe.

Step 3. Name the speakers and download the transcript

Lines appear as they are transcribed - a 30-minute meeting is typically readable within 3 to 5 minutes, each prefixed Speaker 1, Speaker 2. When the full transcript appears, press Name speakers: each voice is listed with how long it spoke and its first line, which for a meeting you attended is enough to place everyone. The names apply across the whole transcript at once.

Then take the text in whichever form the work requires:

• Copy - one click copies the full labelled transcript to your clipboard, ready to paste into meeting minutes, a shared document, or a Jira ticket.
• .txt download - one line per speaking turn, name prefixed. Attach it to the project record or include it in a follow-up email. Opens in any text editor, no proprietary format.
• .vtt download - WebVTT with precise timestamps on every cue. If you keep the .mp4, this file makes it a captioned video: upload both to SharePoint or OneDrive and the captions play in the Teams media player. The subtitles guide covers where the file plugs in on YouTube and other platforms.

Why not use Teams' built-in transcript?

Teams can generate a transcript automatically during a meeting if your IT admin has enabled it and your licence includes the feature. If both of those are true, use it - it is convenient and the output exports as .vtt or .docx directly from the meeting recap. But there are several cases that built-in transcription does not cover:

• Licence gaps. Live transcription in Teams requires Microsoft 365 Business Standard, Business Premium, E3 or E5 at minimum. Business Basic and the free Teams plan do not include it. Many organisations have a mix of licence tiers; not every participant or organiser has access.
• IT has not enabled it. Even on a qualifying plan, transcription must be switched on in the Teams admin centre under Meetings → Meeting policies → Transcription. If your IT team has not done that, the option simply does not appear for anyone.
• The meeting already happened. Teams transcription only works for meetings going forward. A folder of last quarter's .mp4 recordings cannot be retroactively transcribed through Teams - you need something that accepts file uploads.
• Speaker attribution is weak. Teams' built-in transcript labels speakers if they are identified in your organisation directory and are in the meeting from a named account. Guest participants and external callers are often labelled Unknown speaker. Diarisation from an uploaded file attributes by voice regardless of whether the participant had a Teams account.

The upload route covers all four cases. If your organisation's plan already includes Teams transcription and it is working reliably, you do not need this workflow. If not - or for archive recordings - uploading the .mp4 gives you a labelled transcript that matches or exceeds what the built-in feature produces for the cases it handles well.

The adjacent option to both is a meeting-assistant bot (Otter, Fireflies) that joins every call. That is a subscription product with a visible bot participant, and the WhoSayAh vs Otter.ai comparison covers the trade-offs in detail. The short version: a bot makes sense if you want every meeting captured automatically; file upload makes sense if you want control over what gets sent anywhere.

Recording quality: what to fix before you press Record

Transcript quality is mostly decided during the meeting, not after. Three habits pay for themselves:

• Headsets beat laptop microphones. A built-in mic captures fan noise, keyboard clatter and room echo alongside the voice. A headset or USB microphone captures the voice alone. One participant on a bad mic degrades only their own lines - but if that person is the one making decisions, those are the lines you needed most.
• Mute when not speaking, and take turns. Overlapping speech is the hardest case for both word accuracy and speaker attribution. Even 2 seconds of two voices at once can shift the following line to the wrong speaker. Teams' standard mute etiquette is exactly the discipline transcription rewards.
• Start recording before the first agenda item. Decisions have a habit of being made in the first 90 seconds while someone is still saying they will "start the recording now".

What happens to your file

The uploaded file is deleted as soon as the transcript is ready. The transcript itself is deleted after 24 hours. There is no account, so nothing links either to you. The privacy policy states exactly what is stored and for how long; the terms of use list the size and length limits.

For meetings that cannot leave your own network - client calls under NDA, healthcare discussions, legal depositions, any recording subject to a data-residency regulation - the alternative is a self-hosted installation where the audio stays inside your own environment. FastoCloud provides the infrastructure for that: the processing runs inside your own servers, there is no per-minute charge, and the upload workflow for whoever produces the transcript is identical to what is described above.

Teams calls that are recorded for compliance archiving - recorded automatically to a compliance recording policy endpoint rather than to OneDrive - follow a different path entirely. Those recordings are controlled by your compliance team and are not generally downloadable by individual users. This guide covers the standard meeting recording flow, not the compliance-recording flow.

Consent: the conversation to have before you press Record

A Teams recording captures other people. In the EU, the UK, Canada and roughly half of US states, all-party consent is legally required; everywhere else it is still the professional norm. Teams helps by displaying a recording indicator to all participants and, in some configurations, a consent prompt - but do not lean on the indicator alone. Say it out loud before the first agenda item: "I'm recording this for the minutes - any objection?" and stop if there is one. For formal contexts, note the consent in the minutes themselves.

The same principle applies to the transcript: it quotes colleagues and clients verbatim. Share the meeting notes, not necessarily the raw speaker-labelled transcript, unless you know everyone is comfortable with verbatim attribution. If your calls happen on Google Meet rather than Teams, the Meet recording and transcription guide covers that platform's version of this workflow, including a free browser extension that records when Meet itself will not.

Frequently asked questions

Can I transcribe a Teams recording without a Microsoft 365 subscription?

Yes, as long as you have the .mp4 file. If your organisation records meetings to OneDrive or SharePoint, download the file and upload it here. You do not need any Microsoft licence to use WhoSayAh, and there is no account required. The only constraints are the file size and length limits per upload.

What formats does WhoSayAh accept from Teams?

Teams recordings export as .mp4, which is accepted directly. If you extract the audio first, .m4a, .mp3, .wav and .webm all work. Video containers do not need to be converted - only the audio track is read. The size limit is 100 MB and the length limit is 60 minutes per file.

How accurate are the speaker labels for Teams meetings?

Accuracy depends mainly on recording quality and how carefully you set the speaker count. A meeting where participants use headsets and take turns clearly produces very accurate attribution. Crosstalk and bad microphones are the main causes of errors. Entering the exact number of speakers - rather than using auto-detect - is the single adjustment that most improves who-said-what accuracy.

Can I transcribe a recorded Teams Live Event?

Yes. A Teams Live Event records to the organiser's OneDrive as an .mp4, exactly like a regular meeting. Download it and upload it here. Live events often have a single primary speaker with a Q&A panel, so setting the speaker count accurately matters more than usual for large sessions.

Transcribe a Teams recording