Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseAPI / Product · ActionableHabr AI→ original

Anthropic launches Claude Meetings: an assistant in Google Meet that builds a plan as the call happens

Anthropic has launched Claude Meetings, an assistant that joins Google Meet as a separate participant and builds notes, documents, and boards during the conversation. You can ask it questions by voice or in chat, it replies in writing, and the team edits the results during the call. The beta is rolling out on Team and Enterprise plans. So before the meeting ends, ask it for four items: decisions, owners, deadlines, and open questions.

AI-processed from Habr AI; edited by Hamidun News
Anthropic launches Claude Meetings: an assistant in Google Meet that builds a plan as the call happens
Source: Habr AI. Collage: Hamidun News.
Listen to article
Article Fact Sheet & Key Figures
ParameterDetails
Primary sourceHabr AI ↗
Editor-in-Chief
Key metric #1Documents and boards are free until February 10, 2027; tools that connect to other services consume plan limits
Key metric #2Claude stores the transcript and summary, but not the audio recording
Key metric #3Only Google Meet is supported for now, and participants must consent to transcription

Most calls end the same way: everyone hangs up, and an hour later nobody remembers who took what and by when. Usually the unlucky one writes up the outcomes, from memory. On November 12, Anthropic proposed a different approach: Claude Meetings, an assistant that joins a video meeting and builds work materials while people talk.

What Claude does during a meeting

The assistant joins Google Meet as a separate participant. During the call, you can ask it questions by voice or in chat. By default, it answers in writing so it doesn't interrupt the conversation. You can ask it to prepare a document, a board, or slides. Those materials can be edited together with invited colleagues from your organization. If you grant permission, Claude also accesses connected services.

An example from Anthropic's announcement: a team is planning a client event. Claude assembles a document as the discussion unfolds, finds last year's budget in the connected tools on request, and helps move the deadlines onto a shared board. People keep arguing while simultaneously editing the finished plan.

After the call, the meeting page in Claude keeps the notes, tasks, transcript, and created materials. You can return to past discussions and ask questions about them. To connect the assistant, open the Meetings section in Claude, link Google Calendar, and add Claude to the chosen meeting.

Beta terms and pricing

The beta is rolling out gradually on Team and Enterprise plans. Notes, transcripts, and answers that don't require connected tools are included in the plan. Documents and boards created in meetings are free until February 10, 2027. Working with connected tools uses up plan limits. These are the beta terms as of November 13, and they may change after February.

For teams, this means something simple: if Claude goes into the task tracker, storage, and email on every call, limits will run down faster than in regular chat. It makes sense to decide in advance which services to connect to meetings and to check usage across a few trial calls.

The limitations matter too:

  • only Google Meet is supported;
  • Claude saves the transcript, summary, and screen descriptions, but not the audio recording;
  • participants must consent to transcription before the assistant is connected;
  • for client meetings, external guests must be allowed in the settings. On such calls, Claude takes notes but does not reply in the general Meet chat, and external participants do not receive the meeting page. The plan can be shown to the client through screen sharing.

How to avoid a beautifully formatted misunderstanding

The author of a review on Habr, a team lead, sees the new feature as a working tool. The meeting immediately turns into a document, a board, and a task list, so colleagues discuss a concrete plan rather than a recap. There is an extra benefit for those who work with foreign clients. When agreements appear in text right away, it is easier to check whether both sides understood the deadlines and scope the same way, and to fix the wording during the meeting itself.

He suggests asking the assistant for four items: what was decided, who is doing it, by when, and what remains open. That way, during the call, you can see where there is no owner, where the deadlines don't match, and which question was forgotten.

The main risk is an automatic protocol that assigns obligations to people on its own. The phrase "we could try" is not the same as "I'll take it on." So before the meeting ends, participants should confirm out loud the decisions, owners, and deadlines. Otherwise the document will contain a neatly recorded misunderstanding, and there will be no one to hold accountable, since Claude can't take offense.

There is also a human side. When your name is next to a task, the phrase "we were just discussing it" no longer works. But a camera and activity on a call prove nothing by themselves; what should be evaluated is accepted tasks and results. If you adopt such an assistant, it's better to agree up front that it is a team tool, not a way to monitor employees.

A practical sequence: start using Claude Meetings on internal project meetings first. Warn participants about transcription in advance. At the end of the call, read back the outcomes against the four items, and watch limit usage when you connect external services.

Story

Key Questions Answered

What Claude does during a meeting?

The assistant joins Google Meet as a separate participant. During the call, you can ask it questions by voice or in chat. By default, it answers in writing so it doesn't interrupt the conversation. You can ask it to prepare a document, a board, or slides. Those materials can be edited together with invited colleagues from your organization. If you grant permission, Claude also accesses connected services. An example from Anthropic's announcement: a team is planning a client event.

What are the key details on Beta terms and pricing?

The beta is rolling out gradually on Team and Enterprise plans. Notes, transcripts, and answers that don't require connected tools are included in the plan. Documents and boards created in meetings are free until February 10, 2027. Working with connected tools uses up plan limits. These are the beta terms as of November 13, and they may change after February.

How to avoid a beautifully formatted misunderstanding?

The author of a review on Habr, a team lead, sees the new feature as a working tool. The meeting immediately turns into a document, a board, and a task list, so colleagues discuss a concrete plan rather than a recap. There is an extra benefit for those who work with foreign clients. When agreements appear in text right away, it is easier to check whether both sides understood the deadlines and scope the same way, and to fix the wording during the meeting itself.

ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Private AI Builders Community · Vibe Club

Ship AI Faster with Founders, Engineers & Weekly Architecture Reviews

Inside Vibe Club (hamidun.com/club) — weekly live masterminds with Zhemal Khamidun, Tech Radar, full LMS access, and 370+ production Claude Code skills.

Join Vibe Club →Explore Hamidun Academy
What do you think?
Loading comments…