GitHubMIT

AI meeting notes that capture everything

Stop taking notes during meetings. screenpipe transcribes, summarizes, and extracts action items from every call. Works with Zoom, Meet, Teams.

Recording
Q4 Planning Call
Sarah
You
Mike
Sarah:
“The deadline is moved to March 15th...”
screenpipe capturing everything

Quick answer

Automatic AI meeting notes that capture everything. screenpipe transcribes Zoom, Meet, and Teams calls locally. Get summaries, action items, and full transcripts.

See it in action

Meetings are broken

You're either participating or taking notes. You can't do both well.

01

Taking notes means missing what's being said

02

Meeting recordings are too long to rewatch

03

Action items get lost in the conversation

04

Different people remember different things

05

Paid transcription tools charge per minute

100% automatic meeting documentation

screenpipe runs locally and captures all audio from your meetings. Full transcripts, AI summaries, action items - without paying per minute.

Universal compatibility

Works with Zoom, Meet, Teams, Slack Huddles, Discord, or any app with audio.

AI summaries

2-paragraph summary of any meeting. Key decisions and outcomes extracted automatically.

Action item extraction

AI identifies who committed to what. Never lose track of follow-ups.

Searchable history

Ask 'what did Sarah say about the deadline?' and get the exact quote with timestamp.

How it works

1

Enable audio capture

screenpipe captures system audio automatically. No meeting bot needed.

2

Join your meeting

Use Zoom, Meet, Teams - whatever you normally use. screenpipe works in the background.

3

Get your notes

After the meeting, get transcript, summary, and action items.

Key benefits

Be fully present in meetings
Never miss an action item
Search across all past meetings
No per-minute costs
Works offline and keeps data private

Frequently asked questions

Yes, screenpipe captures system audio, which works with Zoom, Google Meet, Microsoft Teams, Slack Huddles, Discord, and any other app that plays audio through your computer.

Transcription accuracy depends on audio quality and the speech-to-text model used. With clear audio and a good model (Whisper large-v3), expect 95%+ accuracy for English. screenpipe supports multiple transcription backends.

Yes, screenpipe supports speaker diarization - identifying who said what. This works especially well when participants have distinct voices or when combined with video analysis.

screenpipe runs locally - no data sent to cloud servers. No per-minute charges for transcription. You can transcribe unlimited meetings without worrying about privacy.

Never take meeting notes again

Let AI handle documentation. You handle the conversation.