Cut audio without re-encoding it
Set a start and an end, and keep what is between them. For an MP3 or an AAC file the cut is a straight copy — no encoder runs, so the part you keep is exactly the audio you started with.
Trim audio
Drag the two handles. Everything outside them is discarded.
Starting the converter…
A cut does not have to cost anything
Trimming is the one edit that can be free. Compressed audio is stored as a long run of small independent frames, so keeping a stretch of them and discarding the rest requires no decoding at all — the frames you keep are copied straight into the new file, untouched.
That is what happens here when the output format matches the input: an MP3 trimmed to MP3, an M4A trimmed to M4A. Nothing is re-encoded, the audio you keep is the audio you had, and the job finishes in about a second whether the source is three minutes or three hours long. Change format at the same time — trim an M4A and take MP3 out — and it becomes an ordinary conversion, with the ordinary cost.
Because nothing is re-encoded, the cut lands on the nearest frame boundary rather than the exact sample you asked for. For MP3 that is about a fortieth of a second. For trimming a lecture, a song or a voice memo it is invisible. If you need a sample-exact cut — lining up a loop, say — convert to WAV in the same pass and the cut becomes exact.
What people actually trim
- Dead air at the start and end. A recording that begins with thirty seconds of a room and someone finding the record button. Trimming this first is also the cheapest way to make a file smaller, because the bytes go away completely rather than being compressed harder.
- The part worth keeping. Ten minutes of a two-hour meeting; the answer to one question in an interview; a single track pulled out of a concert recording.
- A ringtone. Thirty seconds of a song. There is a page specifically for that, because iOS has rules about length and format that are worth handling properly.
- Something too long to send. Trimming beats compressing when there is material you do not need. Cut first, then shrink what is left if it is still too big.
Trimming as the cheapest compression there is
This is worth spelling out, because people reach for a bitrate slider first. Cutting a recording in half halves the file, and the half you keep is completely unchanged. Halving the bitrate also halves the file, but every second of it is now worse.
| Approach | Result | What it costs |
|---|---|---|
| Trim 60 minutes to 30 | Half the size | The half you cut. The rest is untouched |
| Re-encode 192 to 96 kbps | Half the size | Detail across the whole recording |
| Fold stereo to mono | Roughly half again | The stereo image. Free on a single speaker |
Do them in that order and you rarely need the third. For a spoken-word recording all three together take a file to about a quarter of its size with nothing anyone would notice.
What it will trim
Anything the converter reads: MP3, M4A, WAV, FLAC, AAC, OGG, Opus, WMA, AIFF, ALAC, AMR and more, plus the soundtrack inside a video file. The copy-rather-than-re-encode route applies where the codec is already legal in the output container — MP3 to MP3, AAC to M4A or to an .aac, ALAC to ALAC, PCM to WAV. Everywhere else the trim is exact and the audio is re-encoded, which the panel tells you before you run it.
What this page does not do
It trims to a single range. It does not split one file into several parts in one pass, and it does not detect and remove silence automatically. Both are jobs we would rather not half-implement — running the tool twice for two sections works, and it is honest about what it is doing.
It also has the same ceiling as everything else here: a browser tab has roughly two gigabytes of address space, so this warns above 300 MB and refuses above 2 GB.
Step by step
Load the recording
The panel shows its real length, codec and bitrate, read out of the file itself.
Set the start and the end
Two sliders, with a running readout of where you are and how long the result will be.
Choose what comes out
Keep the same format and the cut is a copy with nothing re-encoded. Change format and it is a normal conversion.
Download
You get the section you marked, and nothing else.
About trimming audio
Does trimming reduce the quality of the part I keep?
Not when the output format matches the input. The frames you keep are copied across without being decoded, so that section is exactly the audio you started with. Changing format in the same pass makes it a normal conversion instead.
Learn moreHow accurate is the cut?
Within about a fortieth of a second for MP3, because a copy lands on the nearest compressed frame rather than the exact sample. That is invisible for speech, music or a ringtone. Converting to WAV in the same pass makes it sample-exact.
Learn moreCan I split a file into several pieces?
Not in one pass. Trim to the first section and download it, then set the handles for the second. It is a few more clicks than a dedicated splitter, and we would rather do one thing properly than advertise a mode that half works.
Learn moreIs trimming a good way to make a file smaller?
It is the best way, and the one people skip. Cutting the dead air removes those bytes entirely, while lowering the bitrate degrades every second that remains. Trim first, then reduce the bitrate only if it is still too big.
Learn moreHow long can the file be?
A copy-based trim handles a three-hour recording in about a second, because nothing is decoded. The real limit is memory: a browser tab has roughly two gigabytes of address space, so this warns above 300 MB and refuses above 2 GB.
Learn more