Live meetings: audio, transcript, and layout

Starting and stopping

Choose a project when it matters for context. Start recording when you are ready; stop when the conversation is finished. When you stop, Bearbits can save and continue processing depending on your workflow.

If you leave the page while a meeting is active, the app may warn you. Confirm if you intend to leave so you do not lose a long recording unintentionally.

Microphone and system audio

Microphone captures your voice. System audio captures the sound from your meeting platform (for example, when you share a tab with audio). Together, they help you hear both sides of the conversation in the transcript.

If your browser asks you to share a tab, window, or screen so audio can be included, that share is used only for live transcription. Bearbits does not save a video recording or screen recording from that share, and it does not store the meeting audio as an audio recording.

If others sound quiet or missing, check that system audio is enabled and that your browser is sharing the correct tab or window with audio. If you change audio devices mid-call, you may need to share system audio again.

Transcript and speakers

The transcript updates as you speak. Speaker labels help you see who spoke when; the exact layout depends on how audio is captured.

You can copy the transcript from the meeting view when you need to paste it elsewhere, and you can stop the meeting at any time when you want capture to end.

Language hints and meeting terms

Set language hints so the recognizer matches the languages you expect. You can adjust during a meeting if speakers switch languages.

Meeting terms (glossary) improve recognition of names, acronyms, and domain jargon. Add terms under Advanced → Transcription, combine them with project or global defaults from Settings, and manage project glossary on the project page.

Advanced options (overview)

Many controls live under Advanced options on the meeting toolbar. Open it when you need microphone selection, language hints, glossary terms, AI formatting toggles, system audio, panel visibility, or optional transcript testing.

Defaults matter: for example, including system audio is on by default in Advanced options; include speaker labels in AI and use project context for summary and suggestions are on by default. You can change these when you need a different behavior.

Transcription tab

Pick the microphone before you start recording when possible. Set language hints for the languages you expect in the meeting.

You can add glossary terms and choose whether to include project and global default terms. These are sent to the live transcription engine to improve recognition.

AI Formatting tab

Include speaker labels in AI calls when you want the model to see the full transcript labeling. When this is off, diarization-specific speaker numbers are stripped but AI calls and cleaner transcript copies still keep role labels such as Me and Others.

Use project context pack for summary and suggestions when you want project background included in AI outputs; this is on by default.

Audio tab

Include system audio when you need participants from the meeting app in the transcript. Turn it off if you only want microphone capture.

Panels tab

On small screens, Show Assistant customizations is off by default. Turn it on to reveal additional Assistant controls in the mobile layout.

Testing tab

Transcript testing is optional. When enabled, you can mark an end point on a transcript row to limit what the AI sees for suggestions and summary—useful for careful prompt tuning.