SeaMeet Desktop is here — Record everything, miss nothing. Download free →

How to transcribe a meeting for free

TL;DR

Three genuinely free paths: (1) SeaMeet — 15 minutes of live transcription free with no signup, works in your browser, 20+ languages; (2) Whisper — open-source local transcription, runs on your laptop, unlimited length but requires setup; (3) each meeting platform's built-in transcript (Zoom Cloud, Meet, Teams) — free only on paid plans, which most people already have. All three produce a real transcript. SeaMeet is the fastest to start; Whisper is the most flexible for long recordings you already have on disk.

Try in browser (no signup)

No signup · 15 min free · works on Chrome, Edge, Safari

By SeaMeet Team · Published July 3, 2026

Free tier — 15 minutes, no signup

Method 1 · SeaMeet browser recorder (fastest to start)

The fastest path if you don't already have a recording file. Runs in Chrome, Edge, or Safari on desktop. Captures audio from your browser tab or system, transcribes in 20+ languages live, generates a summary at the end. Free for the first 15 minutes per session — after that transcription pauses (recording keeps going); Sync Pro removes the cap. No account, no email required for the free tier.

  1. 01

    Open app.seameet.ai in a browser tab

    Chrome, Edge, or Safari on desktop (mobile browsers don't expose tab-audio capture).

  2. 02

    Click "Start recording"

    Pick "A browser tab" and select the tab where your meeting is running (Zoom Web, Meet, Teams Web, WhatsApp Web, etc). Tick "Share tab audio" — this is the transcription source.

  3. 03

    Live transcript streams in the panel

    Every word your speakers are playing gets transcribed in real time. Speaker attribution is best-effort based on voice pattern; you can rename speakers after.

  4. 04

    Hit Stop when done

    Recording, transcript, and AI summary appear in your SeaMeet library. Free tier lets you export the transcript to text or clipboard.

Free forever, requires 15-min setup

Method 2 · Whisper (open-source, run locally, unlimited length)

OpenAI's Whisper is an open-source speech-to-text model released in 2022 and improved several times since. You install it on your laptop and feed it audio or video files. It runs offline (no cloud, no data leaving your machine), handles files of any length, supports 90+ languages, and produces a clean transcript. The tradeoff: you need to know how to install a Python package (or use a wrapper like whisper.cpp for a native binary).

  1. 01

    Install whisper via pip or brew

    Simplest: `pip install openai-whisper` (needs Python 3.8+ and ffmpeg on your PATH). For a native binary with no Python, use whisper.cpp: `brew install whisper-cpp` on macOS.

  2. 02

    Grab your meeting recording file

    This works on any recording you already have — Zoom Cloud Recording download, phone screen recording, QuickTime capture, whatever. mp4, mov, mp3, m4a, wav all work.

  3. 03

    Run: whisper recording.mp4 --model medium --language en

    The medium model is a good balance of accuracy vs. speed. Use `--model large` for best quality on important recordings; `--model small` for speed on cheap laptops.

  4. 04

    Read the .txt / .vtt / .srt files it produces

    Whisper writes a plaintext transcript, a WebVTT subtitle file, and an SRT subtitle file to the same folder. Import any into your notes tool.

Whisper doesn't generate a summary — it's a raw transcriber. If you want summaries too, feed the .txt output to ChatGPT or Claude with a prompt like "summarize this meeting transcript into decisions + action items."

Free on paid plans (you may already have it)

Method 3 · Your meeting platform's built-in transcript

Zoom, Google Meet, and Microsoft Teams all generate transcripts automatically as part of their native recording feature on paid plans (Zoom Pro+, Workspace Business Standard+, M365 Business Basic+). If your workspace already pays for one of these, the transcript is a checkbox — no additional tool needed. Free plans on any of the three don't include transcripts.

  1. 01

    Confirm your workspace has the paid tier

    Zoom → check billing page for Pro/Business/Enterprise. Google Workspace → check plan (must be Business Standard+). Microsoft 365 → Business Basic or higher.

  2. 02

    Start the meeting's native recording

    Zoom: Record → Record to the Cloud. Meet: Activities → Recording → Start recording. Teams: More (…) → Record and transcribe → Start recording.

  3. 03

    End the meeting, wait for processing

    Zoom Cloud recording + transcript takes 5-30 minutes. Meet transcripts arrive in the organizer's Drive as a Google Doc. Teams transcripts land in OneDrive + Stream.

  4. 04

    Download or share the transcript

    Zoom: Recordings tab → transcript.vtt / transcript.txt. Meet: Docs → Meet Recordings folder. Teams: OneDrive → Recordings folder.

If you're not the host, or your workspace admin has disabled the feature, this path is unavailable — fall back to Method 1 or Method 2.
Try SeaMeet in browser (no signup) →

15 min free · Chrome, Edge, or Safari on desktop

Platform-specific recording guides

How to record a Zoom meeting How to record a Google Meet How to record a Microsoft Teams meeting How to record a WhatsApp call How to record a Discord call How to record a Slack huddle

FAQ

Frequently asked questions

What's the best free transcription tool if I already have a recording? +
Whisper (Method 2) — it's free, offline, unlimited length, and works on files you already have. SeaMeet's "Import file" flow also transcribes existing recordings; the free tier is limited to 15 minutes of transcript output per import. For short files, SeaMeet is faster; for long files (hour-plus interviews, all-day workshop recordings), Whisper wins.
What accuracy can I expect from a free tool? +
Whisper large-v3 is currently state-of-the-art for open-source; on clean English it reaches ~95% word-level accuracy — better than most paid services. SeaMeet's live transcription is comparable on English and strong on Mandarin/Cantonese/Japanese/Korean/Spanish. Both degrade on heavy background noise, accents, or overlapping speakers.
Are there any real gotchas with free transcription? +
Three to know: (1) two-party-consent jurisdictions require every meeting participant to explicitly consent to being recorded/transcribed — the tool being free doesn't change that; (2) Whisper needs ~2-4 GB of disk for the model files and 4-10 GB of RAM for the medium/large models; (3) SeaMeet's free tier caps at 15 minutes per session and doesn't include cloud sync.
Can I transcribe a meeting I didn't record? +
Not retroactively — transcription needs the audio. If someone else recorded, ask them for the file (mp4, mp3, wav all work with Whisper and SeaMeet). If it's on the meeting platform's cloud (Zoom, Meet, Teams), the recording may already come with a transcript automatically.
How long does Whisper take? +
On modern hardware (M1/M2 Mac, 4090 GPU, or recent Intel/AMD desktop): roughly the same length as the audio for the medium model, ~4× real-time for the large model. On older laptops without a GPU: 4-10× audio length for medium, painfully slow for large. whisper.cpp with GPU acceleration is faster than the Python version.
Does SeaMeet's free tier actually work with no signup? +
Yes for the browser recorder — open app.seameet.ai, hit Start, no email or account. The 15-minute cap applies per session; you can start a new session to reset it. Sync Pro (paid) removes the cap, adds cloud sync, and unlocks longer summaries.
What about "free trial" tools like Otter or Fireflies? +
Their free tiers are usually 300 monthly transcription minutes with a bot participant visible in every meeting. Fine if you can accept the bot; not "free forever" — most users hit the monthly cap fast. See [how to record a meeting without a bot](/en/how-to-record-a-meeting-without-a-bot/) for the tradeoffs.
What's the best free transcription for non-English meetings? +
Whisper handles 90+ languages well — French, Spanish, German, Mandarin, Japanese, Korean, Arabic, Hindi all get near-native accuracy on the large model. SeaMeet's live transcription supports 20+ languages including strong Asian-language support (Mandarin, Cantonese, Japanese, Korean, Vietnamese, Thai, Bahasa).

Ready to try SeaMeet?

No signup. Free for the first 15 minutes. Works in your browser.

Try in browser (no signup) →

Chrome, Edge, or Safari on desktop