> ## Documentation Index
> Fetch the complete documentation index at: https://docs.waterr.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Google Meet

> Send your AI agent into a Google Meet call as a participant.

Paste a `meet.google.com` link and your agent asks to join. Once the host lets it in, it runs your scenario like any other participant: it listens, answers when it is spoken to, and writes up the call when it ends.

## Before you start

* A Google Meet link, for example `https://meet.google.com/abc-defg-hij`.
* A plan that includes meeting platforms (**Starter** and above).
* Someone in the call who can admit guests.

## What participants see

**1. Ask to join**

The agent joins with its persona's name and asks to join. The host gets the usual **Admit** prompt.

**2. In the call**

The agent's tile shows the persona's photo and name. It greets the room and then follows the scenario.

**3. Leaving**

The agent leaves when the scenario ends, when the time limit is reached, or when someone removes it. The meeting carries on without it.

## Good to know

* **The host decides.** If nobody admits the agent, it gives up after a while and the call is marked as not joined.
* **Meet may label the agent.** Google can show guests like ours with a note that they may be automated. That is expected: the agent's name says it is an AI agent.
* **One agent per call is the safe default.** If one of your agents is already in a Meet, a second one may be refused.
* **Some Workspace meetings block outside guests.** If the organizer only allows people from their organization, the agent cannot join.

See [What works where](https://docs.waterr.ai/integrations/overview.md#what-works-where) for the full feature list.


This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.