Skip to main content

Audio trimmer online

Cut audio clips to the exact length you need. Zoom, preview, drag handles, and extract multiple regions.

Free & unlimited
All processing happens in your browser — files never leave your device. The audio engine loads once from a CDN.

Cut a section out of an audio file by dragging handles on a real waveform drawn from the decoded audio, not a generic picture. You can zoom in for sample-level precision, type exact timestamps down to the millisecond, or use the keyboard to nudge the edges. The important detail is what happens on export: if you do not ask for a fade, the cut is performed with a stream copy, so the audio is never decoded and re-encoded and the output is bit-identical to the source in the range you kept. Ask for a fade at the cuts and the file is re-encoded, because a fade has to change the samples. Multi-region mode lets you mark several ranges and export each as its own clip in one pass.

Key facts about Audio trimmer online

Key facts about Audio trimmer online
Lossless pathWith no fade the trim runs as an ffmpeg stream copy — no decode, no re-encode, no quality change
Re-encode pathEnabling a fade forces a re-encode: MP3 and AAC at 192 kbps, OGG and Opus at 192 kbps, WAV to PCM, FLAC to FLAC
Output formatAlways the input's own extension, with a -trimmed suffix on the filename
WaveformDrawn from the decoded buffer with zoom, so you can see the exact transient you are cutting on
Timestamp entryStart and end can be typed as m:ss.mmm for millisecond-accurate edits
PresetsFirst 30 s, last 30 s, first half, second half, middle 60 percent, and remove a 5 s intro
Multi-regionMark several ranges and export them as separate clips named -clip1, -clip2 and so on
Fade controlOne fade length applies to both the in and out point of each exported clip
Accepted inputMP3, WAV, OGG, FLAC, AAC and M4A, up to 200 MB
PreviewPlayback is scoped to the selection, so you hear exactly what will be exported before you export it

What happens to your file

The file stays in your browser. Waveform drawing uses the browser's own decodeAudioData on a copy held in the tab's memory, and the cut itself is performed by ffmpeg compiled to WebAssembly running in a Web Worker on this device. Nothing is uploaded, nothing is queued, and no account exists. The one network request is the first-time download of the ffmpeg core (about 32 MB) from the jsDelivr CDN, a fixed public asset cached by your browser and reused across every audio tool here; your audio plays no part in it.

About this tool

  1. 1

    Load the audio

    Drop an MP3, WAV, OGG, FLAC, AAC or M4A up to 200 MB. The waveform is drawn from the decoded audio as soon as the file is read.

  2. 2

    Set the range

    Drag the start and end handles, or type exact timestamps. Zoom in when you need to land on a specific transient; the waveform redraws at the zoomed resolution.

  3. 3

    Preview the selection

    Playback is limited to the range you marked, so what you hear is what you will get. Adjust the handles while it plays.

  4. 4

    Decide about fades

    Leave the fade off to keep the export lossless via stream copy. Turn it on, and set a length, when a hard cut would click or when you want a clean entrance.

  5. 5

    Add more regions if you need clips

    Switch on multi-region mode and mark each section. Every region is exported as its own file, numbered in order.

  6. 6

    Export and download

    Each clip comes back in the source format with a -trimmed or -clipN suffix, ready to download individually.

Specs & compatibility
Input formatsMP3, WAV, OGG, FLAC, AAC, M4A
Maximum file size200 MB
OutputSame container and extension as the input
PrecisionMillisecond timestamps; the actual cut point on a lossy stream-copy lands on the nearest frame boundary
Clips per passMulti-region mode exports every marked region sequentially, one ffmpeg run each
EngineSingle-threaded ffmpeg.wasm 0.12.10, ~32 MB core fetched once and cached
SpeedA no-fade stream copy is near-instant regardless of length, because nothing is decoded
BrowsersChrome, Edge, Firefox and Safari 15+
  • Keep the fade off whenever you can. Stream copy is both instant and lossless — the exported section is the same bytes the source held.
  • On MP3 and AAC a stream copy can only cut on a frame boundary, so the real cut point may differ from your typed timestamp by up to about 26 milliseconds. Use a fade or a WAV source if that matters.
  • A very short fade, around 10 to 30 milliseconds, is the standard fix for a click at a hard cut and is inaudible as a fade.
  • Zoom before you place a handle on a drum hit or a word boundary. At full-file zoom one pixel can be a second of audio.
  • Use multi-region mode to chop a long interview into topic clips in one pass instead of re-trimming the same file repeatedly.
  • The preset buttons are the fastest way to lop off a known intro or grab a sample from the middle of a track.
  • If you want to remove silence rather than a specific range, the silence remover finds the gaps for you.
  • Zoomable waveform drawn from the real decoded audio
  • Lossless stream-copy export when no fade is applied
  • Millisecond timestamp entry for start and end
  • Six range presets including first and last 30 seconds
  • Multi-region mode exporting several clips per pass
  • Optional fade in and out at the cut points
  • Selection-scoped playback preview
  • Keyboard nudging of the selection edges
  • Cutting a ringtone or a sample out of a song without re-encoding it
  • Removing the first minute of small talk from a recorded meeting
  • Splitting a long DJ set into individual track clips in one session
  • Trimming a podcast intro and outro before publishing
  • Grabbing a quote from an interview to post as a short clip
  • Cutting a voice memo down to the part you actually need to send
Not if you leave the fade off. With no fade the tool asks ffmpeg to stream-copy the selected range, meaning the compressed data is copied straight through without ever being decoded. The output is bit-identical to the corresponding part of the source. Switch a fade on and the file must be re-encoded, because changing the volume envelope means changing the samples — that costs one generation on a lossy source. WAV and FLAC are lossless either way.
Because a stream copy cannot cut inside a compressed frame. MP3 stores audio in frames of 1152 samples, about 26 milliseconds at 44.1 kHz, and AAC works similarly. When the trim is lossless, the cut lands on the nearest frame boundary rather than your exact timestamp. For most editing that is invisible. When you need the cut exactly where you typed it, either enable a fade — which forces a re-encode and therefore sample-accurate cutting — or work from a WAV.
Yes. Turn on multi-region mode and mark each range you want. On export the tool runs the trim once per region and returns each as its own file, suffixed -clip1, -clip2 and so on, in the original format. Each clip gets the same fade setting. This is the efficient way to break a long recording into segments, because the file is only loaded once and each cut is a separate fast stream copy when no fade is in play.
MP3, WAV, OGG, FLAC, AAC and M4A, up to 200 MB. The output always matches the input: trimming an MP3 gives you an MP3, trimming a FLAC gives you a FLAC. There is no format conversion in this tool by design, because keeping the container identical is what makes the lossless stream-copy path possible. If you need a different format afterwards, send the trimmed file through the audio converter as a separate step.
It is drawn from the real decoded audio, sampled down to the pixel width of the canvas, and it redraws at higher resolution as you zoom in. That means what you see genuinely corresponds to the amplitude at that point in the file, so you can place a cut on a drum hit or between two words by eye. Zoom matters: at full-file view a single pixel can represent a second of audio, which is far too coarse for a precise edit.
No. The waveform comes from decoding the file in your own tab, and the cut is performed by ffmpeg compiled to WebAssembly running in a Web Worker on your device. The file never leaves the browser. The only network request is the one-time download of the ffmpeg engine itself, about 32 MB of fixed public code from the jsDelivr CDN, which your browser caches and reuses for every audio tool on this site. It is identical for every visitor and carries no part of your file.

Related tools

View all

Part of Audio tools that run in your browser

Updated

We use anonymous analytics to improve ToolChamp. No personal data is stored or sold. Privacy Policy