Audio Cutter

It selects one continuous segment of audio from your input file and writes that segment to a new file in the same format. Drop in a 60-minute MP3, mark a selection from 12:30 to 13:05, click Cut, and you get a 35-second MP3 back. The output file inherits the codec, sample rate, bit depth, and channel count of the input. No conversion, no re-mastering, no automatic level adjustment — just the slice you asked for. If you want to convert formats at the same time, run the cut first, then put the result through one of the format conversion tools.

Cutting always re-encodes the audio. This is so cuts can land at any sample, not just at codec-frame boundaries (which would force you to round to the nearest 26ms or so for MP3, or the nearest container-frame for AAC). The trade-off depends on the source format. Lossless inputs (WAV, FLAC, AIFF) re-encode to a bit-identical lossless output — every sample in the selection is preserved exactly. Lossy inputs (MP3, AAC, OGG, Opus, M4A) get one transcoding pass at the original bitrate, which introduces a small amount of additional quality loss that is essentially inaudible on typical playback gear at 192 kbps and above. If quality is critical and you have the lossless source, cut from that. For 99% of use cases (ringtones, podcast clips, voice memos), the transcoding loss is invisible.

The other online audio cutters work by uploading your file to a server, processing it there, and streaming the result back. We do not. The browser downloads a copy of FFmpeg compiled to WebAssembly (cached after the first visit), reads your file from local memory, runs the cut, and writes the result to a Blob you download. No upload step. No server-side temporary storage. No logs of what you cut. This matters for client recordings, NDA-covered material, leaked tracks, voice memos with personal details, sensitive interviews, or anything else you would rather not hand to a third-party server. Disconnect from the internet after the page loads and the tool still works.

You can technically cut audio by typing in HH:MM:SS timestamps blindly, and the tool does support that for precision work. But for most jobs, the waveform is the difference between three guesses and one cut. The visual makes it obvious where the speaker stops talking, where the song's chorus starts, where the silence between sections lives, and where the drum hit lands. Drag the handle to that visual landmark, preview, adjust by a fraction of a second if needed, and you are done. Typing 00:01:23.450 by hand only beats the waveform when you already know the exact timestamp from a transcript.

This cutter is built for the common case: one file, one segment, one download. If you need to cut the same file in 15 different places, batch-process a folder of recordings, edit at the level of individual MP3 frames without re-encoding, or apply effects like fades and crossfades, a desktop editor like Audacity (free) or Reaper (paid) is the right tool. They have a learning curve this tool deliberately avoids, but they pay off on heavy lifting. For everything else — and especially for one-off cuts — you do not need the install.