Google Meet integration
Give MeetStream a Google Meet link and a bot joins the call to record, stream, and transcribe it. The bot shows up as a participant, so the room always knows it is there.
What this integration gives you
- Send a bot into any Google Meet call from its link with one API call
- Per-participant audio and video streams across the whole meeting
- MP4 recordings and real-time media over secure WebSocket at roughly 200ms latency
- Real-time and async transcripts with speaker identification
- Send messages, images, and animated GIFs into the Meet chat
- No host permission or paid Workspace plan needed to join
Joining a Google Meet call
MeetStream joins Google Meet from the meeting link. You call the API with the link and a display name, and the bot enters the call as a visible participant. Because Google Meet often gates external attendees, an admit prompt may appear for the host, who lets the bot in just as they would any guest knocking to join.
There is nothing to install on the organizer's Google account and no Workspace tier requirement. The bot operates from the link, which keeps the integration simple: your product does not need Google admin access to start capturing meetings, only the link to the call your user wants recorded.
Capturing Meet audio, video, and transcripts
On Google Meet you get separate streams per participant for both audio and video. That granularity feeds accurate speaker identification and lets you build layouts or analysis that treat each speaker independently rather than working from a single blended recording.
Media is available live over a secure WebSocket at roughly 200ms latency for real-time features, and as an MP4 recording once the meeting ends. Transcripts are delivered both in real time and as a final async version, each labeled by speaker. The bot can also post into the Meet chat, including images and animated GIFs, which is useful for confirmations, prompts, or agent responses inside the call.
- Per-participant audio and video for every attendee
- Live media over WebSocket for real-time processing
- MP4 recording delivered after the meeting
- Speaker-identified transcripts, live and final
How to connect
- Generate an API key from the dashboard at app.meetstream.ai.
- Call the create bot endpoint with the Google Meet link and a bot display name.
- Select capture options: recording, real-time streaming, and transcription.
- Point a webhook at your endpoint to receive recording ready and transcript ready events.
$ curl -X POST \
https://api.meetstream.ai/api/v1/bot \
-H "Authorization: Bearer $API_KEY" \
-d '{"url": "zoom.us/j/74648392"}'
Frequently asked questions
Can the bot join a Google Meet it was not invited to?
The bot joins from the meeting link. Google Meet may prompt the host to admit external participants, so the organizer clicks to let the bot in, the same way they admit any guest. You do not need Workspace admin access.
Are Google Meet participants captured separately?
Yes. MeetStream provides per-participant audio and video on Google Meet. Each attendee is a distinct stream, which improves speaker identification and supports per-speaker processing downstream.
Can I stream a Google Meet in real time?
Yes. Live audio and video are available over a secure WebSocket at roughly 200ms latency, so you can transcribe, coach, or run agents while the meeting is still happening rather than waiting for the recording.
Where is the detailed Google Meet API reference?
This page covers connecting and capturing. For endpoints, parameters, and payloads, see the Google Meet product page and the docs at docs.meetstream.ai.
Connect it in an afternoon
One API to join, record, stream, and transcribe meetings across Zoom, Google Meet, and Microsoft Teams.