Extract key points, decisions, and action items from meeting transcripts. Automatically identifies participants and tasks.
This tool runs 100% in your browser. All computation happens locally on your device — your input is never uploaded to any server. Results are for reference only.
The meeting summarizer turns a raw transcript into structured meeting minutes through a three-stage pipeline. First, the raw text is diarized — every line is attributed to a speaker by matching name prefixes, so the engine can count participants and per-speaker word shares. Next, decisions and action items are extracted using keyword and pattern rules such as "we will" or "assigned to", which the local SmolLM2 model can refine into natural-language summaries in AI mode. Finally, everything is assembled into a structured result: an executive summary, a decisions list, and action items with assignees and priorities. Because the entire pipeline runs in your browser — including optional Whisper transcription of meeting audio — no meeting content ever reaches a server.
A three-stage pipeline: raw transcript is diarized per speaker, key phrases are classified into decisions and tasks, and a structured summary is assembled for export.
Sarah pastes a 40-line transcript from her team's weekly product sync, with speaker labels such as "Mia: I'll own the onboarding revamp." The engine diarizes three speakers, flags two decisions, and produces four action items with assignees and priorities.
To summarize a meeting: paste your meeting notes or transcript — the AI extracts key decisions, action items, and discussion points, all processed locally in your browser.
FreeToolHub AI Meeting Summary is a free browser-based tool that generates meeting summaries and extracts action items from notes, no signup.
Turn meeting transcripts into summaries. Extract participants, decisions, and action items with priority. 100% local, no upload, free.
The AI Meeting Summarizer converts long meeting transcripts into the few things anyone actually rereads: who attended, what was decided, and what happens next. You can paste a transcript directly, or upload an audio recording (MP3, WAV, or M4A) that is transcribed on your device in more than fifteen languages, with automatic speaker labeling that separates who said what. The output is structured rather than a wall of text: a concise summary, a list of decisions, and action items that each carry an assignee, a priority of high, medium, or low, and a deadline when one was stated. Everything exports in the formats teams already use.
Project managers who sit in six meetings a day use it to keep a reliable record of commitments without writing minutes by hand. Remote and distributed teams summarize calls across time zones so absent members catch up from the source, not secondhand retellings. Agency and client-service staff turn discovery and check-in calls into shared next-step lists. Executives and board secretaries produce consistent minutes from governance meetings where decisions must be documented precisely. Students and researchers summarize group discussions and interviews. Because transcripts and audio are processed locally, it also works for HR conversations, medical case reviews, and any recorded discussion too sensitive for a cloud transcription service.
(1) Provide input: paste a transcript and an optional meeting title, or upload an audio file, choose its language, and let local Whisper transcription with speaker diarization produce the text, with stage-by-stage progress while it runs. (2) Click Summarize. With a local AI model loaded, the tool writes natural-language output; without one, it falls back instantly to keyword and pattern detection and tells you which engine ran. (3) Review the results: participants, summary, decisions, and action items with owners and priorities. (4) Export as Markdown, Notion-friendly Markdown with the transcript in a toggle block, Google Docs HTML with an action-item table, plain text, or an iCalendar file that creates a calendar event carrying the summary.
For transcripts, any pasted text works: Zoom, Teams, or Google Meet exports, court-reporter-style text, or notes typed live. Speakers can be marked inline, for example "Sarah: let's ship on Friday", and the tool detects participants from those patterns. For audio, upload MP3, WAV, or M4A recordings and pick from more than fifteen transcription languages, including Chinese, Spanish, French, German, and Japanese. Diarization then labels speakers automatically and reports speaking statistics per person. Files are processed entirely on your machine: the Whisper model runs in your browser alongside audio decoding, so an hour-long internal recording never touches a third-party server at any stage.
No. The AI model runs 100% in your browser using WebGPU acceleration. Your meeting transcript is tokenized, processed, and summarized locally on your device. No text is transmitted over the network, making it safe for confidential board meetings, legal depositions, and HR discussions that cannot leave your machine.
It accepts plain-text transcripts from Zoom (export from the recording portal), Microsoft Teams (download from meeting chat), Google Meet (email transcript), or any platform providing text. Paste directly or upload .txt files. Transcripts up to 30,000 characters (roughly a 45-minute meeting) are processed in a single pass.
Output contains four sections: a 3-5 sentence executive overview, key decisions made (with context), action items mapped to assignees with deadlines, and a ready-to-send follow-up email draft. The entire summary generates in 10-30 seconds depending on transcript length and your GPU capability.
This tool is also known by these tasks — each link opens the same tool with a focused guide:
What do you call a crab that plays baseball?
No paywalls, no signups, no data sold. Built by a solo developer who believes useful tools should be accessible to everyone.
☕Support me on Ko-fi— keep tools free100% of proceeds go towards hosting & building more free tools.