AI Meeting Notes

Turn a meeting transcript into a summary, decisions and action items - free, private, no upload.

AI meeting notes turn a wall of transcript text into something a team can actually act on: a short summary, the decisions that were made, and the action items that came out of them. Paste a transcript below - the free version runs entirely in your browser, with nothing uploaded.

0 words

The free engine never uploads your transcript - everything runs in this browser tab.

Engine

Not live yet. When it ships, this option will send your transcript to a larger model on Meetrix's servers for better accuracy - clearly disclosed before it runs. The free local engine above is unaffected.

Model size

Works best on clear, English meeting transcripts. Strong accents, crosstalk or a non-English transcript will make the notes noticeably weaker - that's a limit of a small model, not a bug.

Why use this meeting notes generator

Runs in your browser

The free option summarizes your transcript locally, in a Web Worker. Nothing is uploaded.

Decisions and action items, separated

Not just a summary - a separate, editable list of what was decided and what still needs to happen.

Edit before you export

Fix a missed detail or add an owner to an action item directly, instead of rewriting the notes from scratch.

How to generate AI meeting notes

  1. Paste a meeting transcript above, upload a .txt/.srt/.vtt file, or click Load a sample transcript if you don't have one handy.
  2. Leave the model size on Balanced, or pick Fast for a quicker pass or Best quality for a slower, more thorough one.
  3. Click Generate Meeting Notes. The first run downloads the model once; this can take a minute depending on your connection.
  4. Watch the progress bar while the transcript is summarized in parts, then organized into sections.
  5. Read the Summary, Decisions and Action Items - click into any of them to fix wording or add something the model missed.
  6. Copy the notes or download them as .txt or .md.

Common AI meeting notes problems and fixes

  • The first run is slow: that's the one-time model download, not a bug. It's cached afterwards, so every later run on this device starts immediately.
  • The summary is too generic: a known limit of small models on long or dense transcripts. Try the Best quality model size, or edit the summary directly to add specifics.
  • An action item or decision got missed: click + Add decision or + Add action item and add it yourself - editing is expected to be part of the workflow, not a fallback.
  • Notes repeat the same point in two sections: small models sometimes duplicate a point across the summary and a list. Delete the extra copy - the underlying content is still correct.
  • No text comes back at all, or it's very short: check the transcript actually has content and isn't just timestamps or speaker labels with no dialogue - the tool needs real spoken content to summarize.
  • Transcript is too long and won't generate: split it into two or three parts at natural breaks (a break, a topic change) and generate notes for each separately.
  • It won't run in this browser: the local engine needs WebAssembly, and runs faster with WebGPU. Update to the latest Chrome, Edge, Firefox or Safari.
  • I only have an audio or video recording, not a transcript: run it through the audio to text tool first - its output pastes directly into this tool.

Need a transcript first? Use the audio to text tool to turn a recording into text, then paste the result in here. Once you know what a meeting is costing in people's time, the meeting cost calculator puts a number on it. Browse all Meetrix features, or see how Meetrix captures notes automatically during a live call on the Meetrix blog.

Frequently Asked Questions

Is this AI meeting notes generator free?

Yes. The local engine is free to use, with no sign-up and no limit on the number of transcripts - just the per-transcript length cap listed above.

Is my transcript uploaded anywhere?

Not on the free engine. Your transcript is summarized entirely in your browser, in a background worker, and never leaves your device. The only path that uploads anything is the optional high-accuracy server engine, which discloses the upload before it runs and is not live yet.

What's the difference between the free and high-accuracy options?

The free option runs a small instruction-tuned model locally in your browser: no upload, no cost, but weaker than a large model on nuance, sarcasm, and transcripts with a lot of crosstalk or filler. The high-accuracy option (coming soon) sends your transcript to a larger model on Meetrix's servers for noticeably better notes, but requires an upload and an email address.

How long can my transcript be?

Up to about 10,000 words on the free local engine - past that, a browser tab doing this many chunks of AI generation gets slow and memory-hungry. Longer transcripts: split them into parts and generate notes for each, or wait for the high-accuracy option.

Do I need to record or transcribe a meeting first?

Yes - this tool turns an existing transcript into notes, it doesn't listen to audio directly. If you only have a recording, run it through the audio to text tool first, then paste the result in here.

Can it tell who said what?

Only as well as your transcript already does. If your transcript has speaker names (like "Alex: ..."), the model can pick up on them when assigning action items. A transcript with no speaker labels at all will produce notes with no ownership information either.

Why are the notes generic or missing something important?

Small models summarize well but aren't as sharp as a large one at catching every detail, especially in a long or noisy transcript. Edit the notes directly to add what's missing, try the Best quality model size, or wait for the high-accuracy server option.

Does it work on non-English transcripts?

Not well. The local model is tuned mostly on English text, so notes from a transcript in another language will be noticeably weaker or may not follow the expected format at all.

How long does this take?

Anywhere from a few seconds for a short transcript to a few minutes for a long one on an older machine - the transcript is split into parts and summarized one at a time, then combined into the final notes. The progress bar shows which part is being processed.

Can I edit the notes before exporting?

Yes. Click into the summary or any decision or action item and type - you can also add or remove list items. Every export (copy, .txt, .md) uses your edited version.

What formats can I export?

Plain text (.txt), Markdown (.md) with checkboxes on action items, or copy straight to your clipboard.

Which browsers are supported?

Any recent Chrome, Edge, Firefox or Safari. Chrome and Edge can use your GPU (WebGPU) and run faster; other browsers fall back to WebAssembly, which works everywhere but is slower.

Want this generated automatically, live, during the call?

Meetrix records meetings and generates AI notes automatically as part of the call, on your own self-hosted infrastructure - no separate transcript-and-paste step required. See Meetrix's self-hosted AWS products.

Talk to Us