Free, unlimited, no account
Edit Audio by Editing Text
Drop in a podcast, interview or voice memo, get a word-timed transcript, then delete words to cut the recording. Video files work exactly the same way. Transcription runs on our servers for accuracy and every language; the editing and export happen right in this browser tab, and there is nothing to sign up for.
Transcription runs on our servers, so there is no model to download — it starts right away and auto-detects the language. A Chromium browser on a desktop is recommended.
Need clean speaker separation or publish-grade accuracy?
This editor transcribes on our servers with Whisper — fast, and it auto-detects the language. When you need clean speaker separation, diarized multi-speaker labels or accuracy you can publish, transcribe it on ConvertAudioToText instead, export SRT or VTT, and import that transcript here to do the cutting.
How it works
Four steps. Transcription runs on our servers; the cutting and export run right in your browser — no queue, no job to wait on.
1. Drop in your recording
Audio or video: MP3, WAV, M4A, MP4, WebM, MOV, MKV. The browser reads it off your disk and extracts the audio, unless you sent it over from your CATT dashboard.
2. It transcribes on our servers
The audio is sent to our servers and transcribed with Cloudflare's Whisper (large-v3-turbo), which detects the spoken language automatically across ~99 languages and returns a timestamp on every single word. The audio is auto-deleted within a day. If our servers are ever unavailable, it falls back to a smaller model that runs privately in your browser.
3. Delete words, the audio cuts
Select text, press backspace, and the matching stretch of the recording is cut. Playback skips the cuts live so you hear the edit before you export.
4. Export the final cut
ffmpeg.wasm trims and re-joins the kept ranges in the browser and hands you an M4A (or MP4 if you brought video) to download.
What people use it for
Free podcast editor in the browser
Trim tangents, kill dead air, tighten an intro, export an M4A. Speaker labels make multi-guest episodes readable at a glance.
Remove filler words from a recording
One click cuts every um, uh, ah, eh, hm and mm in the transcript, matched against a fixed list of English filler sounds. It is a literal word match rather than a judgement call, so like, so and right stay in. The button only appears when the spoken language is set to English, because the same list lands on real words elsewhere: um is German for in order to and Portuguese for a. Undo puts everything back.
Cut video by deleting text
Video works exactly like audio. No timeline scrubbing: read the transcript, delete the sentence you do not want, and the cut lands on the word boundary.
Your audio isn't kept
Audio is uploaded only to transcribe it — on our servers with Cloudflare Whisper — then auto-deleted within a day. We don't store it and there is no long-term retention. The editing and export happen in your browser, and recent projects live only in your browser's IndexedDB, which you can clear at any time. If our servers are ever unavailable, transcription falls back to running entirely on your device.
We do record anonymous product analytics on our own self-hosted PostHog: page views and four events telling us the tool loaded, a transcription started, an export finished, or a link out was clicked. No file names, no audio, no transcript text.
What it needs from your machine
- A Chromium desktop browser is recommended. The page is served cross-origin isolated so SharedArrayBuffer is available for multi-threaded WebAssembly.
- WebGPU is used when your machine exposes it; otherwise it falls back to WebAssembly, which works but is slower.
- Everything is held in RAM, so very long recordings can exhaust the tab. Short clips are where this tool is happiest.
Frequently asked questions
Is the editor really free?
Yes. There is no account, no sign-up, no export limit, no watermark and no paid tier of this editor. Transcription runs on our servers and we cover that cost; you pay nothing.
Does my audio get uploaded anywhere?
The audio does, to transcribe it. It is uploaded to our storage, transcribed on our servers with Cloudflare Whisper, and auto-deleted within a day — we do not keep it and there is no long-term storage. Everything else stays local: the video/audio is decoded in your browser with ffmpeg.wasm, the cutting happens on your machine, and the export is assembled in the browser. If our servers are ever unavailable, transcription falls back to a model that runs entirely in your browser, where nothing leaves the device at all. We also record anonymous product analytics that contain no file data.
Does it work with video too?
Yes, and identically. Drop in an MP4, WebM, MOV or MKV and you get the same word-timed transcript, the same delete-to-cut editing and an MP4 export with the picture intact. The tool is built around the transcript, so audio is the shortest path to a result, but nothing about video is second class.
Is a speech model downloaded to my browser?
Normally no. Transcription runs on our servers, so nothing large is downloaded. A model is only fetched if our servers are unavailable and the editor falls back to on-device transcription — Whisper Base is roughly 200 MB, pulled from the Hugging Face CDN once and cached — so the tool keeps working offline in that rare case.
Which browser should I use?
A recent Chromium browser (Chrome, Edge, Brave, Arc) is the recommendation. The editor needs SharedArrayBuffer, which requires the cross-origin isolation headers this site sends, and it uses WebGPU when your machine exposes it, falling back to WebAssembly otherwise. Firefox works but is usually slower. Safari and most mobile browsers are not reliable for this workload.
How long a file can I edit?
Transcription handles long recordings — the audio is chunked and processed on our servers, so length is not a limit there. The editing and export run in your browser and hold the media in memory, though, so a very long file can still strain a laptop tab. For a marathon two-hour interview you can also transcribe it at ConvertAudioToText, download the SRT, and import it here just to do the cutting.
Does it work in languages other than English?
Yes. The server transcription auto-detects the spoken language across ~99 languages, so you do not need to pick it — say what you said and it comes back in that language. (In the rare on-device fallback, the smaller model cannot auto-detect, so it uses the language you pick or English.) Filler removal is the one English-only part: the list is a literal match on English filler sounds, and outside English it lands on words that carry meaning, so the button stays hidden unless the language is English.
Can I bring my own transcript?
Yes. Pick Import transcript on the source menu and load an SRT, VTT or JSON caption file. Skipping Whisper means no model download and no waiting, and it is the fastest path for long recordings.
How accurate is the transcript?
Whisper Base and Small are the two models small enough to run in a browser, so they trade accuracy for size. They are good on clean speech and get noticeably worse with accents, crosstalk, background noise or non-English audio. For a transcript you intend to publish, use a larger model on ConvertAudioToText and import the result.
What formats can I export?
Audio-only projects export as M4A. Video exports as MP4 (H.264 video, AAC audio). Cuts are re-encoded rather than copied, so they land exactly on the word boundary instead of the nearest keyframe.