Audio Trimmer

1. 01Choose an audio fileDrag it in or click to pick one. The tool reads its length and fills in the full range, so the end time starts out at the end of the file.
2. 02Set the start and endType times as seconds, m:ss or h:mm:ss, with hundredths if you need them — 5.25, 0:05.25 and 1:05.75 are all accepted. The selection length updates as you type.
3. 03Preview the selectionPlay just the part you have chosen. Playback starts at your start time and stops at your end time, so you hear the clip before making it.
4. 04Trim and downloadTrim the file, check the length of the result, and download it. Start over at any point to work on a different file.

• Pull one quote out of a recordingTake the thirty seconds you actually need out of an hour-long interview or lecture recording.
• Cut the silence off a voice memoTrim the dead air at the start and the fumbling at the end before sending a recording to someone.
• Make a short clip to shareCut a section of a song or a podcast down to something small enough to send in a message.
• Trim a ringtone or alert soundCut a few seconds out of a longer track to use as a notification sound.

• 0:05.25 to 0:17.80A 12.55-second clip from near the start of a file. Fractional times are respected — you are not limited to whole seconds.
• Drop the first 30 seconds of a podcastSet the start to 0:30 and leave the end where it is to remove an intro.
• One minute out of the middle of a lectureStart 32:10, end 33:10 — the h:mm:ss form handles long recordings without you converting anything to seconds.

• Preview before you trimTrimming takes a moment on a long file. Previewing is instant, and it is the fastest way to find out your start time is a second early.
• Give yourself a little headroomStarting a fraction of a second before the first word and ending a fraction after the last one sounds far more natural than cutting exactly on them.
• Convert an odd file firstIf a file will not open here, converting it to WAV or MP3 first and trimming the result almost always works.
• Trimming an MP3 twice costs qualityEach trim re-encodes. Trimming a clip you already trimmed here is fine for speech, but go back to the original file if you can.

Whatever you put in, what comes out is an MP3 at 192 kbps. That is a deliberate choice, and it has a cost worth stating plainly: the audio is decoded and encoded again rather than copied, so this is not a lossless cut, and a WAV or FLAC file will come back as a lossy MP3. The alternative — copying the original stream untouched — can only cut at frame boundaries, which would quietly move your cut by up to a few hundredths of a second while the screen showed the time you typed. One honest, proven encoder that cuts where you asked is better than a lossless promise that silently moves the edit. For a single trim of speech or music at 192 kbps the loss is not something most people will hear; for archival work, keep your original.

The cut is made after decoding, not by seeking the container, so it lands on the time you typed rather than on the nearest frame. What is not exact is the length of the finished file. MP3 stores audio in frames of about 26 milliseconds that cannot be split, and the encoder adds a small amount of padding of its own, so a clip asked for at 12.55 seconds typically comes back within a few hundredths of a second of that rather than at exactly that. The result panel shows both numbers — what you asked for and what the finished file measures — so you can see the difference instead of taking our word for it. This is not a sample-accurate editor, and it does not claim to be.

The trimming is done by ffmpeg compiled to WebAssembly, running inside the page. Your audio is read by the browser, handed to that engine in memory, and written back out in memory — there is no request that carries the file, no server-side processing, and no copy kept anywhere. The one thing that is fetched from the network is the engine itself: two files, a small loader and the WebAssembly binary, downloaded the first time you trim something and then cached by your browser. Those requests carry no audio. The file name and the times you set are not sent either; analytics only see that the tool was used.

Files up to 100 MB are accepted. Audio processing needs more memory than the file size itself: your file is read into memory once, copied again into the engine’s own in-memory filesystem, and the engine and the finished clip sit alongside both — all inside a single browser tab. The limit is deliberately conservative rather than tuned to a number, because the same page has to work on a phone, where that headroom is a fraction of a desktop’s and the failure mode is the tab being killed outright rather than running slowly. 100 MB still covers a 90-minute podcast at 128 kbps; past it, the honest thing is to refuse up front.

There is no waveform, no multi-track view and no fade handles — it is a range and two time fields. That is enough for cutting one section out of one file, which is the job this tool is for. It is also not a converter: if what you need is to turn a video into audio, or change an audio file from one format to another without cutting it, Convert to MP3 does that.

How do I trim an audio file online?

Open the file here, type a start and end time, preview the selection to check it, then trim and download. Nothing is installed and nothing is uploaded — the file is trimmed by your own browser.

Which audio formats can I open?

MP3, WAV, M4A, AAC, OGG, Opus, FLAC and WebM audio. That is a tested list rather than a claim to handle everything: these are the formats the engine reliably decodes in a browser. Video files are not accepted here — use Convert to MP3 to extract the audio first.

What format is the trimmed file?

Always MP3, at 192 kbps, whatever you started with. The tool re-encodes rather than copying the original stream, so a WAV or FLAC input comes back as a lossy MP3.

Is the trim lossless?

No. The clip is decoded and encoded again, so some quality is lost — for most speech and music at 192 kbps it is not noticeable, but it is a real re-encode and this tool will not pretend otherwise. Keep your original file if you need the audio untouched.

How accurate is the cut?

The cut is made at the time you type, not at the nearest frame boundary. The finished file’s duration can differ from your selection by a few hundredths of a second, because MP3 frames are about 26 ms long and the encoder adds a little padding. The result panel shows both the length you asked for and the length the finished file actually measures.

Can I use fractions of a second?

Yes. Times take hundredths — 5.25, 0:05.25 and 1:05.75 all work, as do plain seconds and h:mm:ss for long recordings.

Is my audio uploaded anywhere?

No. The file is read and trimmed inside your browser. The only network requests the tool makes are for the WebAssembly audio engine itself, which is downloaded once and then cached; those requests do not contain your audio.

How large a file can I trim?

Up to 100 MB. Audio processing needs more memory than the file size itself, so the limit is kept conservative to avoid browser crashes, especially on phones.