> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.meetstream.ai/welcome/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.meetstream.ai/_mcp/server.

# MeetStream

> MeetStream is a meeting bot API: send bots into Zoom, Google Meet, and Microsoft Teams to record, transcribe, and interact in real time through a single integration.

[Quickstart](/guides/get-started/create-your-first-bot)

[How bots work](/guides/introduction/how-bots-work)

[![MeetStream quickstart](/_fern-img/261a54c966836c8dea896a22646e8679171553b396310196742148f86b2a74f8.webp)**Quickstart**One call puts a bot in the meeting. Recording starts the moment it joins.](/guides/get-started/create-your-first-bot)

[![MeetStream API reference](/_fern-img/45b15a7a8cbdf6e403e9d9566f7f9db1ed14fe25eddc9e1b94e6bf9500c07eb6.webp)**API reference**Every endpoint for bots, transcription, calendars and voice agents.](/api-reference/introduction)

[![Build with MeetStream AI tools](/_fern-img/8ff291604fc892fc757018837a42b5c1a84ec361cbd0a7e43541e397b3476baa.webp)**Build with AI**MCP server, CLI and agent skills, so your coding agent writes the integration.](/build-with-ai/meetstream-mcp-server)

## What the API does

#### Media capture and streaming

Export mixed audio, isolated per-participant tracks, or full-meeting video. Stream audio and video live over WebSocket while the call is still running.

#### Transcription and diarization

Speaker-labelled transcripts from the MeetStream engine or five third-party providers, post-call or streamed live over webhook.

#### Agents that act in the call

A bot can speak using text to speech, post in the meeting chat, and display an image as its video feed. MIA agents listen and answer live, with wake words and interruption handling.

#### Scheduling and control

Sync a calendar for auto-dispatch, set auto-leave rules, pause and resume recording mid-meeting, and follow every state change over webhooks.

## What people build

#### Sales coaches

* Bot joins live calls and streams speaker-attributed transcripts over WebSocket
* Your intent layer detects objections and pricing questions from the live stream
* Coaching cues surface to the rep mid-call, while the conversation is still open
* A post-call agent drafts the follow-up and the action items

#### Notetakers

* Bot joins instant or scheduled meetings and transcribes with speaker diarization in real time
* Your pipeline extracts decisions, action items and open questions as structured objects
* Every action item keeps its speaker and source timestamp, so it stays auditable
* Output lands in Notion, Jira or Confluence through their APIs

#### Meeting summarizers

* One API covers Zoom, Google Meet and Microsoft Teams, so there is a single integration to maintain
* A webhook fires the full attributed transcript to your pipeline the moment the meeting ends
* The structured summary goes to Slack, Notion or email
* Raw transcripts stay available through the API for anything downstream

#### CRM automation

* Bot captures the call with speaker attribution
* Your extraction layer pulls deal stage, contact details and notes from the transcript
* Fields map onto CRM objects through their REST API, with no manual entry
* Works with any CRM that exposes an API; the field mapping is yours to configure

> Meeting bot API documentation for Zoom, Google Meet and Microsoft Teams: create bots, stream real-time audio, transcribe, and run in-meeting voice agents.