How transcription works

Every recording is transcribed automatically — no action needed. Open a meeting and switch to the Recording tab to see the transcript alongside the audio player, with each line attributed to a speaker and timestamped.

Turnaround

Transcription starts as soon as the recording ends and typically finishes within a few minutes, faster for shorter meetings. The AI report (see How your report is generated) runs after transcription completes, so a report that hasn't appeared yet usually means transcription is still in progress.

What affects accuracy

• Audio quality at the source — a muted mic, background noise, or a bad connection limits what can be transcribed no matter how good the model is.
• Overlapping speech — people talking over each other is the hardest case for any transcriber, Meetbook included.
• Uncommon terminology — product names, acronyms, and jargon are more likely to be misheard on first pass.
• Language — see Settings → Meeting → Spoken language; leaving it on Auto Detect works well for single-language meetings but can struggle with code-switching.

Speaker labels

New speakers show up as generic labels ("Speaker A", "Speaker B") until you rename them. See A speaker is mislabeled — how to fix it to rename or merge them.