How to Use AI to Summarize Meetings and Auto-Generate Action Items (Free Tools)

The Meeting Notes Problem

Meeting notes are one of those tasks that everyone knows they should do well and almost no one does. Either nobody takes notes, one person takes terrible notes while half-listening, or someone types furiously and produces a transcript that nobody reads.

AI meeting tools solve this by doing three things at once: transcribing the full conversation, identifying who said what, and extracting the summary and action items automatically. You can be fully present in the meeting because the documentation is handled.

For Zoom Meetings: Otter.ai

Otter.ai’s OtterPilot feature joins your Zoom meetings automatically (you give it calendar access and it shows up). During the meeting, it transcribes in real time and identifies speakers. After the meeting, it generates a summary, pulls out action items with the name of the person responsible, and sends it to all participants.

The live transcript is visible during the call, which is useful for participants who want to catch up if they joined late. The free tier gives 300 minutes per month — enough for 10-15 standard meetings.

For Microsoft Teams: Copilot for Microsoft 365

If your organization uses Teams, the Copilot integration is the easiest option. Enable Copilot transcription at the start of a meeting (or in settings for all meetings). After the call, you get a full recap: what was discussed, decisions made, action items with owners. You can also ask Copilot questions about the meeting afterward — ‘what did Sarah say about the Q3 timeline’ — and it searches the transcript.

For Google Meet: Fireflies.ai

Fireflies connects to Google Calendar and joins your Meet calls automatically. The AI analysis includes sentiment tracking, topic highlights, and CRM integration. For teams that use HubSpot, Salesforce, or similar tools, Fireflies can automatically push meeting notes and action items into the CRM record for the relevant contact or deal.

For Offline or Recorded Meetings

Option A: Upload to Otter or Fireflies

Both Otter and Fireflies accept audio/video file uploads on paid plans. Record your in-person meeting on your phone, upload the file, and get a transcript and summary back within minutes.

Option B: Whisper + Claude

The free workflow: record the meeting on your phone. Transcribe with MacWhisper (free, local). Paste the transcript into Claude with this prompt: ‘This is a meeting transcript. Produce: (1) a 5-bullet summary, (2) a list of action items with owners, (3) any open questions or unresolved issues.’

Making Meeting Summaries Actually Useful

The most common failure mode is generating a great meeting summary and then nobody reading it. Build a habit: at the end of every meeting, the AI summary goes into one place — a shared Notion page, a Slack channel, or directly into your project management tool. Fireflies and Otter both have integrations that automate this.

• Zoom or Meet: Use Otter.ai (free tier) for automatic summaries
• Microsoft Teams: Enable Copilot transcription in meeting settings
• Sales calls: Use Fireflies for CRM integration
• In-person meetings: Record on phone, transcribe with Whisper, summarize with Claude
• Share summaries to a consistent location — Slack, Notion, or your PM tool

How AI Meeting Summarizers Work

An AI meeting summarizer transcribes your call, then uses AI to condense it into key points, decisions and action items with owners. Some join your calls live and produce notes automatically; others work from a recording or transcript you provide. The value is turning an hour of conversation into a scannable summary and a clear task list in seconds – work that used to require someone taking manual minutes. Accuracy depends on clean audio and a good underlying transcription model.

Free vs Paid Meeting Summarizers

You can summarise meetings for free by transcribing the audio – for example with a Whisper-based tool – and asking a general AI assistant to extract key points and action items. Dedicated platforms add convenience: automatic call joining, speaker labels, integrations and shared notes, usually on paid tiers with limited free plans. For occasional summaries, the free transcribe-then-summarise route works well; for regular team meetings, a dedicated tool saves real time.

Getting Accurate Summaries and Action Items

Good input gives good output. Record clean audio, and if you use a general assistant, prompt it clearly – ask for decisions, action items with owners, and any open questions, in that structure. Always skim the summary against your memory of the meeting to catch anything misheard, especially names and numbers. A quick review turns a good automatic summary into a reliable record you can act on and share.

What is the best AI meeting summarizer?

For teams, a dedicated tool that joins calls and labels speakers is most convenient; for individuals, transcribing with a free tool and summarising with an AI assistant works well and cheaply. The best pick depends on whether you value automation and integrations or a free, flexible workflow.

Related: the best AI transcription tools, the best Otter.ai alternatives, the best AI podcast tools, and the best AI tools for freelancers.