Compress and convert audio, free
MP3, WAV, M4A and OGG in every direction — to a file size you name or a bitrate you pick, optionally downmixed to mono. All inside your browser, nothing uploaded.
Drop an audio file here, or choose one
MP3, WAV, M4A or OGG, up to 100 MB and 10 minutes. The recording stays in your browser — we do not upload it anywhere.
You can drop several audio files at once — up to ten. They all get the same format and bitrate, and the finished ones come back as a single ZIP.
It works on a phone too, but slower: the processing runs on your own device's processor. A desktop machine is quicker for longer recordings.
When is compressing worth it?
Almost always when a file will not fit somewhere: an e-mail attachment, an upload form, a messenger. Audio is generous material here, because most of the file size can be dropped without anyone hearing it. An hour of speech in WAV is more than half a gigabyte; the same thing in 64 kbps Opus is around thirty megabytes, and stays perfectly intelligible.
That is why there is a "by target size" mode. Instead of dragging a vague quality slider, you say how many megabytes have to fit — and we work the bitrate out from that. For audio this is exact arithmetic rather than trial and error: at a constant bitrate the size follows from the length.
Which format?
MP3 if it has to play anywhere: an old car radio, a memory stick, every program ever written. M4A if it is going to a modern device — AAC audibly outperforms MP3 at the same bitrate, and it is native on an iPhone. OGG (Opus) if size matters most: for speech it stays clean even at 32–64 kbps, which is out of MP3's reach. And WAV when you are not compressing at all but taking the sound into an editor.
One thing is worth knowing: converting between lossy formats always loses a little, because the detail that was thrown away cannot be brought back. Start from the original, best-quality file where you can — not from something already compressed once.
Frequently asked questions
What does downmixing to mono do?
It merges the left and right channels into one. For speech that usually takes nothing away — a dictaphone records the same thing on both sides anyway — and the file ends up roughly half the size. For music it is audible: the stereo image disappears.
Why is the finished file bigger than my target?
Because the recording is long enough that the target could only be met at an unacceptable quality. In that case it is encoded at the lowest sensible bitrate, and we say so in advance. The fix is a shorter section or a bigger target.
Are the cover art and the tags kept?
The embedded cover art is not: most target formats cannot carry it, so we drop it. The title and artist fields usually survive.
Is there a daily limit?
No. The processing runs on your own machine, so it costs us nothing how often you use it — and your files never go anywhere.