convert
Zoom recording to text
Use a completed Zoom recording you are authorized to access. An audio-only export can be easier to handle than a video containing the same conversation.
Last checked:
Tiny transcription · 40.8 MB CPU / 119.7 MB GPU weights. Runtime files download too.
Choose a file or record your microphone.
Download and processing settings
Model weights: 40.8 MB for WebAssembly; 119.7 MB for GPU. Configuration, tokenizer and runtime files add a smaller download. Source: Hugging Face file sizes, checked 2026-10-30.
Model file sizes and source · Last checked October 30, 2026
Tiny uses the least memory. Base and Small may improve recognition but need more memory and time. Try a short sample first. No accuracy or speed is guaranteed.
Audio is processed locally. Model files download from Hugging Face when you start, then stay in your browser cache when storage permits. How privacy works
Your transcript
Click a word or timestamp to listen. Search updates as you type. Edit text and add speaker names yourself. Speaker recognition is not automatic.
Shape your subtitles
SRT and VTT exports use this layout. These controls limit characters and lines, not reading speed. Check subtitle reading speed before publishing.
Saved on this device
Recordings and edits are stored in this browser when local storage is available. Export an important transcript before clearing browser data. The delete control clears app storage and its model cache. Use browser cache controls to remove any browser-managed download copies too.
1
00:00:00,520 --> 00:00:04,180
I mentioned to the universe, sir, I exist.
Get the right recording
Find the completed recording in the location selected by Zoom, or download an authorized cloud recording. Select its audio or video file after conversion finishes.
| Check | Zoom recording workflow |
|---|---|
| Typical content | Audio or video saved by the recording workflow |
| What to preserve | The original file and the final edit used for timestamps |
| Common issue | A meeting transcript should distinguish participant speech from generated meeting summaries. The app does not join a live call. |
Decode before choosing a bigger model
The browser must decode the entire file before speech recognition can begin. Its support depends on the codec, container and operating system. A filename extension alone cannot guarantee success. MurmurScribe reports a decoding error if this browser cannot read the selected recording.
Test playback and a short representative passage first. Choose the spoken language if you know it. Use transcription to keep the original language, or translation when an English draft is the intended result. Tiny is the lightest starting model. A larger model may need substantially more memory and time.
Review and export
A meeting transcript should distinguish participant speech from generated meeting summaries. The app does not join a live call. After transcription, listen to the timestamps near any uncertain sentence. Correct proper names and technical terms manually. Speaker names are manual tags; the app does not infer identities from voices.
Choose TXT for plain notes, DOCX for an editable document, or JSON for structured segment data. For captions, export SRT or VTT and test it against the same media file. Subtitles need sensible timing and line breaks, not just a correct transcript.
Your audio is processed locally. The model download is separate from the recording and can be reused from browser cache. Saved transcripts are tied to this browser profile, so download a copy before clearing local storage or moving to another device.
Sources
- Browser audio codec guide · MDN · Checked 2026-10-30
- Media containers and browser support · MDN · Checked 2026-10-30
- Browser decoding and resampling · MDN · Checked 2026-10-30
- Zoom computer recordings · Zoom · Checked 2026-10-30