Meeting transcription
What is meeting transcription?
A transcript makes a call searchable and quotable, and it is the raw material an AI meeting summary is written from.
On this page
How does meeting transcription work?
A speech recognition model listens to the audio in short chunks and writes down the words it hears, with a timestamp for each line. Before it can write anything, it has to know the language; modern systems detect it from the audio itself, and some can switch when a different person speaks another language.
The other half is knowing who said what. When a tool transcribes a single mixed recording, it has to guess where one voice ends and the next begins. A meeting service that receives each participant's audio separately can attach names far more reliably. Transcripts can be produced live, line by line during the call, or after the meeting from the full audio.
What affects a transcript's accuracy and use?
| Factor | Effect |
|---|---|
| Microphone and room | A close microphone and a quiet room give the clearest text; noise and echo cause dropped or wrong words. |
| People talking over each other | Overlapping speech is the most common source of errors. |
| Names and jargon | Unusual names, product terms and abbreviations are often misspelt. |
| Language switching | Works best when each speaker stays in one language per sentence. |
A transcript is also a record of what people said. Tell participants before it starts, follow the consent rules that apply to you, and decide who should receive it and how long it should be kept.
How does transcription work in Meeto?
In Meeto, transcription is part of AI Summary, available to Pro and Team hosts. The host selects “Start AI Summary” and chooses who receives it afterwards, “Only me” or “Everyone in the call”. Everyone in the call sees a banner and an indicator while it runs. It captures speech, not video. Starting a recording also starts AI Summary.
The transcript names each speaker, the language is detected per speaker so multilingual meetings work, and the summary is written in the meeting's main language. During the call there is a live transcript; afterwards the meeting page has Summary and Transcript tabs with key points, decisions and action items, a transcript search, and export to Markdown, text or a PDF summary, plus a share link the owner can switch off.
Transcripts and summaries are kept until you delete them, while recorded video is kept for 90 days. Meeting content is not used to train AI models. See the AI summary page for the details and noise suppression for one way to get cleaner text.
Frequently asked questions
Is a transcript the same as live captions?
No. Live captions are short lines shown over the video while people speak and then disappear. A transcript is a document of the whole conversation that you can search, export and keep. In Meeto, the live transcript appears in the AI Summary panel during the call and on the meeting page afterwards.
Do participants need to agree to be transcribed?
They need to know, and depending on where you and they are, the law may require their consent. Meeto shows everyone a banner and an indicator when transcription starts; the host is responsible for any consent the situation requires.
What happens after the monthly AI allowance is used up?
Pro and Team include 10 hours of AI summaries a month per account. After that, the full transcript and a basic summary continue until the allowance resets.
Put it to work
Open your Meeto account to set it up, or explore the working product first.