Text-Based Audio Editor: Edit Audio by Editing Text, Online & Free
Delete a word from the text and the audio goes with it. Free, in your browser.
A text-based audio editor turns your recording into text with a timestamp on every word, then cuts the audio when you cut the text. Delete a sentence and the matching stretch of audio goes with it. CATT Editor does this free at edit.convertaudiototext.com, with no account, and every step runs inside your own browser tab.
The editor is a separate free app on edit.convertaudiototext.com. It opens in a new tab, asks for nothing, and reads your file straight off your disk.
Works best in a Chromium desktop browser. The first run downloads the speech model, about 200 MB for Base or about 600 MB for Small, then caches it. Everything is held in memory, so short clips work best.
Why use Text-Based Audio Editor?
Delete text, cut the audio
Select a word or a whole sentence, press backspace, and the matching stretch of media is removed. The cut lands on the word boundary, and playback skips it straight away so you hear the edit before you export it.
Strip the English filler sounds in one click
The editor carries a fixed list of English filler sounds: um, uh, ah, eh, hm, mm, uh-huh and their common spellings. When the transcript holds any of them, a Remove fillers button appears with the count, and one click cuts them all at once. It is a literal word match rather than a judgement call, so like, so and right stay in, and the same list is applied whatever language you spoke: on a non-English recording it can flag a word that carries meaning, which is what undo is for.
Your file never leaves the device
There is no upload and no server-side pipeline. The browser reads the file off your disk, decodes it with WebAssembly, and runs the speech model on your own hardware. We do record anonymous product analytics, and they carry no file names, no audio and no text.
No account, no export limits
No sign-up, no watermark, no credit balance, no paid tier of the editor. What limits you is your own machine, not a plan.
Video files work the same way
Drop in MP4, WebM, MOV or MKV and edit it exactly like audio. Video projects export as MP4 with H.264 video and AAC audio, and audio-only projects export as M4A.
Know the accuracy trade-off
The in-browser model is small enough to run on a laptop, so it is good on clean speech and gets worse with accents, crosstalk, background noise and languages other than English. For text you intend to publish, or a two-hour recording, run it through ConvertAudioToText first and bring the result into the editor.
How it works
Drop in your file
Audio or video: MP3, WAV, M4A, MP4, WebM, MOV or MKV. The browser reads it straight off your disk, so there is no upload, no queue and no progress bar to wait on.
The speech model runs on your machine
Whisper runs in a Web Worker on your own CPU or GPU and returns text with a timestamp on every single word, grouped by speaker. The first run downloads the model, about 200 MB for Base or about 600 MB for Small, then caches it.
Edit the text, the audio follows
Delete the words you do not want. Playback skips the removed ranges immediately, so you hear the result before you commit to it.
Export the final cut
ffmpeg.wasm trims and re-joins the kept ranges in the browser and hands you a file to download: MP4 for video projects, M4A for audio-only ones.
Text-Based Audio Editor vs editors that need an account
What changes when the editing happens on your machine instead of theirs.
| Capability | CATT | Editors that need an account |
|---|---|---|
| Account to start | None | Sign up before the first edit |
| Where your file goes | Stays on your device | Uploaded to their servers |
| Cost to export | Free, no watermark | Free-tier caps, watermarks or both |
| Install | None, one browser tab | Often a desktop app for the full feature set |
| How you cut | Delete the words, the audio follows | Varies by tool |
| Long or difficult recordings | Capped by your browser memory | Processed on their servers |
Frequently asked questions
Yes. There is no account, no sign-up, no export limit, no watermark and no paid tier of the editor. All the work happens on your own machine, so there is no per-minute cost for us to pass on to you.
No. Your browser reads the file locally, decodes it, and runs the speech model on your own hardware. The media, the audio and the text never leave the device. The only network requests the page makes are the app itself, the one-time model download, and anonymous product analytics that contain no file data.
Because the speech model runs on your machine, it has to get there first. Whisper Base is roughly 200 MB and Whisper Small is roughly 600 MB. The browser caches the model after that first run, so transcription and export keep working if your connection drops. Loading the page itself still needs the network: there is no offline app cache.
A recent Chromium browser on a desktop, so Chrome, Edge, Brave or Arc. The editor needs SharedArrayBuffer and 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.
We do not impose a limit, but your browser does. Everything is held in memory, so a laptop is comfortable with clips up to roughly ten to twenty minutes and gets slow or runs out of memory past that. For a two-hour interview, run it through ConvertAudioToText first and bring the text into the editor.
Yes. Pick Import on the source menu and load a timed transcript file you already have. Skipping the on-device model means no download and no waiting, and it is the fastest path for long recordings.
Whisper Base and Small are the two models small enough to run in a browser tab, 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 text you intend to publish, use the larger models on ConvertAudioToText.
Video projects export as MP4 with H.264 video and AAC audio. Audio-only projects export as M4A. Both are written by ffmpeg.wasm inside your browser and saved straight to your disk.
An audio trimmer works on the waveform: you pick a start time and an end time. This works on the words: you read the text and delete what you do not want, and the cut lands on the word boundary. Use the trimmer when you know the timestamps, and this when you know the sentence.
Related tools
Free Descript Alternative
The edit-by-text part of Descript, free in your browser. Not the rest of the suite, and we list what is missing.
Audio Trimmer
Trim and cut audio files to the exact length you need
Podcast Transcription
Turn podcast episodes into searchable, shareable text