Audio Trimmer

Cut a section out of an MP3, M4A, WAV or OGG file on a waveform, add fades, and export a WAV without uploading. 100% free, no signup. Everything runs in your browser.

100% free No sign-up Private by design Works on any device
Audio TrimmerRuns locally

Loading the tool…

Most audio trimming is a thirty second job: the first few seconds of a recording are someone saying "is this on", the end is a chair scraping, and the middle is the bit you want. I built this because the online cutters all want the file uploaded first, and the desktop editors are a hundred megabytes of software for a task that needs two sliders. Drop the file, the browser decodes it and draws the waveform, drag the in and out points, press play to hear exactly the selection, and export.

The decoding uses the browser's own audio engine, which handles MP3, M4A, WAV, OGG and in most browsers FLAC. Everything after that is arithmetic on sample arrays done in the page: the cut, a short fade at each end so the edges do not click, and optional normalisation to bring a quiet recording up to a sensible level. Output is a 16-bit WAV at your device's audio rate, lossless, which every editor, phone and DAW opens. Nothing is uploaded, which matters for a voice memo or a recorded call more than for most files. For capturing audio in the first place, the screen recorder records a tab or the microphone.

How to use

  1. Drop an audio file onto the box, or click to choose it. The waveform appears once it has decoded.
  2. Drag the Start and End sliders. The highlighted region on the waveform is what will be kept.
  3. Press Play selection to hear exactly that region. Adjust and play again until the cut lands where you want.
  4. Leave the fades at their defaults, or lengthen them for a gentle start and finish. Tick Normalise if the recording is quiet.
  5. Press Export WAV, then download. The file is named after the original with a trimmed suffix.

Why use our audio trimmer?

You hear the cut before you commit. Play selection plays precisely the region between the two markers, nothing before and nothing after. Finding the exact word where a clip should start takes two or three tries with the slider, and the waveform makes silence and speech easy to tell apart at a glance.

The edges do not click. A cut that lands mid-wave produces an audible tick at the start or end of the clip. The default fades of twenty milliseconds are too short to hear as fades and long enough to remove the tick, which is the detail that separates a clean trim from one that sounds edited. Longer fades are there for when you want them to be heard.

Lossless out. The export is PCM WAV at your device's audio rate, usually 44.1 or 48 kHz since the browser resamples on decode, built with a correctly sized RIFF header rather than the approximations some quick encoders write, which is why it opens in strict players as well as forgiving ones. Nothing is re-compressed, so a trimmed MP3 loses no quality it had not already lost. If you need an MP3 afterwards, any editor will make one from the WAV. The Web Audio API is what makes decoding in the browser possible.

Normalisation for quiet recordings. Voice memos and phone recordings are often far below full scale. Normalise finds the loudest point in the selection and raises everything so that point sits just under the limit, which makes a quiet file usable without introducing distortion. Silence is left alone rather than amplified into hiss.

Nothing uploaded, and an honest note about size. The file is decoded and cut on your device. A WAV is roughly 10 MB per minute of stereo, so a long clip makes a large file; that is the cost of lossless, and the page tells you the size before you export. For most trims the clip is short and the size is nothing. The create ZIP tool bundles several exports for sending.

Who is this tool for?

Cleaning up a voice memo or interview. Cut the preamble and the trailing noise, normalise so the speaker is audible, and the file is ready to share or transcribe. The recording never leaves your machine, which for an interview with a client or a patient is the point.

A ringtone or notification sound. Twenty seconds of a track, faded in and out, exported as WAV and converted on the phone. The fades matter more here than anywhere, because a ringtone that starts with a click is a ringtone you replace.

Extracting a quote or a clip for a presentation. A sentence from a recorded talk, trimmed tight with a short fade, dropped into a slide. The video to GIF converter does the equivalent for a few seconds of video.

Shortening a recording before it goes anywhere. A two hour meeting recording with ten useful minutes in it. Trim to the ten minutes and the file that gets shared is a tenth of the size and contains only the part that should be shared.

Frequently asked questions

Is the audio file uploaded?

No. The file is decoded by your browser's audio engine and cut in the page. There is no server and nothing is retained. For a recorded call, a voice memo or anything with another person's voice in it, that is the reason to use this rather than an online cutter.

Why is the output a WAV and not an MP3?

WAV is lossless, opens everywhere, and can be produced in the browser without shipping an encoder. Encoding MP3 in the page would mean re-compressing audio that was probably already compressed, losing quality twice. Trim here losslessly, then let an editor or the phone make the MP3 if you need one; the screen recorder page explains the same trade for video.

What formats can I trim?

Whatever your browser can decode: MP3, M4A and AAC, WAV, OGG and Opus, and FLAC in most browsers. If a file fails to decode, the codec is one the browser does not carry, and converting it to WAV or MP3 first will fix it.

What do the fades do?

They ramp the volume from silence at the very start of the clip and back to silence at the very end. At the default twenty milliseconds they are inaudible as fades and exist purely to stop the cut from clicking. Set them to a second or two for a deliberate fade in or out.

What does normalise do?

It finds the loudest sample in the selection and scales the whole clip so that sample sits just under full scale. The result is louder without distortion, which is what a quiet voice memo needs. It does not compress or even out the volume within the clip; it raises everything by the same amount.

How big will the file be?

About 10 MB per minute for stereo at 44.1 kHz, half that for mono. The page shows the estimate before you export. If the result needs to be smaller, trim tighter, or convert the WAV to MP3 afterwards in any editor.

Can I cut out the middle and keep both ends?

Not in one pass; the tool keeps one continuous region. Export the first part, then export the second, and join them in an editor. Keeping the tool to a single cut is what keeps it quick for the common case.

Related tools