Drag across the waveform, hear the selection, download just that part.
🔒 Runs entirely in your browser — nothing you type is uploaded or stored on a server.
Drop in an audio file, drag across the waveform to choose the part you want, listen to check it, and download just that piece. It handles MP3, WAV, M4A, OGG and FLAC, and it will happily take the sound out of a video file too.
Also on Txtset: transcribe or generate speech on your device.
Nothing is uploaded. The file is decoded by your browser’s own audio engine and cut on your device. No queue, no email address, no file-size charge, no ten-minute limit — and no copy of your recording sitting on somebody’s server afterwards. For an interview, a voice note or an unreleased track, that is the difference that matters.
The waveform is drawn from the real samples, one peak per pixel column, so you can see where the silence is and where the speech starts instead of guessing at a timeline. Type exact times into the start and end boxes if you would rather be precise than drag, and add a fade so the cut does not begin and end with a click.
The output is a WAV — uncompressed, written straight from the samples your browser decoded, with no second round of lossy compression. One honest caveat: a browser decodes audio at its own playback sample rate, usually 44.1 or 48 kHz, so a file recorded at some other rate is resampled on the way in. The page tells you the rate it is working at.
How to use
- Drop an audio file on the box, or click to choose one. A video file works too — its sound track is what gets decoded.
- Drag across the waveform to select the part you want to keep. The selection is highlighted and everything else is greyed out.
- Fine-tune with the Start and End boxes if you need an exact time. Both accept seconds, or
m:ss.ss. - Press Play to hear exactly what you have selected before committing to it.
- Add a fade if the cut lands mid-sound, then download the WAV.
Examples
A ringtone
Drag out thirty seconds of a song, add a 0.5 second fade at each end, and download — no app and no upload.
Trimming an interview
The waveform makes the silence before someone starts speaking obvious, so the dead air at the front comes off in one drag.
Sound from a video
Dropping in an MP4 decodes its audio track, so you can pull a quote out of a recording without any video software.
Frequently asked questions
Is my audio uploaded to a server?
No. There is no server behind this page. Your browser decodes the file in memory, the samples are cut on your device and the result is saved locally. You can disconnect from the internet once the page has loaded and it still works.
Why does it only save WAV and not MP3?
Because writing an MP3 needs an encoder, and the honest options are a large library downloaded from a third party or a claim we cannot keep. A WAV is written from the decoded samples directly, so your audio is not squeezed through a lossy codec a second time — which is exactly what trimming an MP3 in most tools does to it. If you need an MP3 at the end, convert the WAV once, rather than re-compressing on every edit.
The WAV is much bigger than my MP3. Is something wrong?
No, that is what uncompressed audio is. A WAV holds every sample at full precision, so a minute of stereo CD-quality audio is about ten megabytes whatever it came from. It is the right format to cut to, and the wrong format to keep forever.
What file types can I open?
Whatever your browser can play: MP3, WAV, M4A and AAC, OGG and Opus, FLAC, and usually the audio track inside MP4, WebM and MOV. If your browser cannot decode it, the page says so rather than failing silently.
How long a file can it handle?
Up to 200 MB, which is roughly three hours of MP3. The whole file is decoded into memory to draw the waveform and cut it precisely, so the ceiling is your device’s memory rather than an arbitrary limit.
What does the fade do?
It ramps the volume up at the start of your selection and down at the end. Cutting mid-waveform leaves a sudden jump in the signal that you hear as a click; a fade of even a tenth of a second removes it.
Can I remove a section from the middle instead?
Not in one pass. Trim the part before it, trim the part after it, and join them in an audio editor. This page cuts one continuous selection, which is what almost everyone actually needs.
Does mixing to mono halve the file?
Yes, near enough, and it is often the right call for speech. Two channels of the same voice carry no extra information; music is a different matter and should stay stereo.
Why is the sample rate different from my original file?
Because a browser decodes all audio at the rate its own sound output runs at — commonly 44,100 or 48,000 Hz — and hands back samples at that rate. There is no way to ask it for the file’s native rate instead. For anything above about 32 kHz this is inaudible, but it is worth knowing if you are working to a spec. The rate actually being used is shown under the waveform.
