MeetStream API Docs
Everything you need to put a bot in a Zoom, Google Meet, or Microsoft Teams call and get recordings, per-participant media, and transcripts back. One API call to your first bot.
Explore the documentation
Quickstart
Create your first bot with one authenticated request and a meeting link.
API reference
Every endpoint on api.meetstream.ai, with parameters, responses, and examples.
Authentication
Create an API key and send it as an Authorization: Token header.
Webhooks and events
Bot lifecycle and processing events. Best effort delivery, so return 2xx fast.
Live transcription
Stream transcript segments during the call with speaker labels and turn signals.
Per-participant audio
Separate audio per speaker instead of one mixed track.
Per-participant video
Separate video streams for each participant across all three platforms.
Calendar integration
Connect a calendar once and let bots auto-join scheduled meetings.
Zoom setup
Zoom App Marketplace configuration for production bots.
MCP server
Give Claude, Cursor, or any MCP client direct access to the API.
Agent skills
Drop-in SKILL.md files so coding agents build on MeetStream correctly.
Migrate from Recall.ai
Map concepts, swap the create-bot call, and move webhooks across.
Start with one request
The whole product is reachable from a single call. POST a meeting link and a bot name to create_bot, and a bot joins the call as a visible, named participant. There is no host-side plan to upgrade and no admin install for your users, and the same call works across Zoom, Google Meet, and Microsoft Teams.
From there you either wait for webhooks or poll for status. When the meeting ends, processing events tell you the recording and the speaker-labeled transcript are ready to fetch.
Machine-readable and agent-ready
The API is published as an OpenAPI 3.0 spec, so you can generate a client, load it into Postman, or hand it to a coding agent. Every documentation page is also available as clean Markdown by appending .md to its URL, and any section exposes a scoped index at /llms.txt.
For agents that should drive the API rather than read about it, the MCP server exposes the bot lifecycle as tools, and the CLI wraps the same endpoints for the terminal.
- OpenAPI 3.0 spec in JSON and YAML
- Markdown version of every docs page
- MCP server for Claude, Cursor, and other clients
- Command line interface for scripting
- SDKs in eight languages
What the API returns
Recordings and MP4 video, separate per-participant audio and video on all three platforms, diarized transcripts with speaker labels, participant join and leave events, in-meeting chat, and screenshots. Real-time audio streams over a secure WebSocket at roughly 200ms, and bots can act inside the call by sending messages, displaying images, or speaking through text to speech.
curl -X POST https://api.meetstream.ai/api/v1/bots/create_bot \
-H "Authorization: Token $MEETSTREAM_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"meeting_link": "https://zoom.us/j/74648392",
"bot_name": "MeetStream Bot"
}'Frequently asked questions
Where is the MeetStream API reference?
The full API reference lives at docs.meetstream.ai/api-reference. It documents every endpoint on api.meetstream.ai, including create_bot, bot status and detail, transcripts, recordings, participants, calendar auto-join, and the in-meeting actions.
How do I authenticate with the MeetStream API?
Every request carries an Authorization header in the form `Token YOUR_API_KEY`. Create a key in the dashboard under the API tab, then send it on all calls to https://api.meetstream.ai.
Is there an OpenAPI spec I can import?
Yes. The OpenAPI 3.0 spec is published at docs.meetstream.ai/openapi.json and docs.meetstream.ai/openapi.yaml, so you can generate a client or load it into Postman or an agent toolchain directly.
How do I send a bot into a meeting?
POST to /api/v1/bots/create_bot with a meeting_link and a bot_name. The response returns identifiers you can poll or wait on, and lifecycle webhooks tell your app when the bot joined, started recording, and finished processing.
Can AI coding agents read these docs directly?
Yes. Append .md to any docs URL for clean Markdown, append /llms.txt to any section for a scoped index, or connect an MCP client to the documentation server for search. The action MCP server at mcp.meetstream.ai drives the API itself.
Build your first bot today
One API call to a live bot, then record, stream, and transcribe across Zoom, Google Meet, and Microsoft Teams.