Convert WAV to MP3
Compress huge WAV recordings to MP3 without an upload. See exactly how much smaller each file gets, with no 100 MB cap and no queue.
- Runs in your browser
- Files are never uploaded
- Free, no account
First use downloads the 10 MB video engine from this site. After that it is cached and works offline. It is WebAssembly served by onthisdevice.com — never a third-party CDN — and your file is still processed entirely on your own device.
The conversion upload services are worst at
WAV is the format that breaks free web converters. A one-hour stereo interview is roughly 600 MB; a multitrack bounce or a field recording can be several gigabytes. Free tiers cap at 100 MB, 500 MB or "2 files per hour" precisely because they have to receive that file, store it, encode it on their hardware and send it back.
None of that applies here. The LAME encoder is compiled into a WebAssembly build of FFmpeg that runs inside this browser tab, so the file never leaves your machine and never enters a queue. The practical ceiling is your own memory — the browser sandbox tops out at roughly 2 GB of working memory, which comfortably covers a three-hour stereo recording.
How much smaller, exactly
Standard CD-rate stereo WAV is 1411 kbps. The saving is simply the ratio of your chosen MP3 bitrate to that number.
| MP3 setting | Share of the WAV | One hour, stereo | One hour, mono |
|---|---|---|---|
| 320 kbps | 22.7% | ≈ 137 MB | ≈ 137 MB |
| 256 kbps | 18.1% | ≈ 110 MB | ≈ 110 MB |
| 192 kbps | 13.6% | ≈ 82 MB | ≈ 82 MB |
| 128 kbps | 9.1% | ≈ 55 MB | ≈ 55 MB |
| 96 kbps | 6.8% | ≈ 41 MB | ≈ 41 MB |
MP3 bitrate is the total for the file, so forcing mono does not halve the size on its own — but it does mean the whole bitrate is spent on one channel, so a mono voice recording at 96 kbps sounds like a stereo one at 192 kbps. For a single-voice interview, switching channels to mono and dropping to 96 kbps is the strongest move available.
Choosing a bitrate by content
- Single-voice interview or dictation: 96 kbps mono. Sounds identical to the WAV over headphones and lands around 40 MB an hour.
- Podcast with two or three speakers: 128 kbps. The platform standard; anything higher is inaudible on a commute.
- Live recording, room ambience, applause: 192 kbps stereo. Diffuse noise and audience sound are what low bitrates handle worst.
- Music mixdown you are sending to someone: 256 or 320 kbps. Also send the WAV if they will be mastering it.
- Anything heading for further editing: do not convert. Keep the WAV; MP3 is a delivery format, not a working one.
When to switch on loudness normalisation
The advanced panel has an EBU R128 loudness normaliser targeting −16 LUFS with a −1.5 dBTP ceiling — the level podcast platforms and most speech distributors expect.
Turn it on for interviews, lectures, meeting recordings and anything captured across several sessions at different gain settings; it is the difference between a listener constantly reaching for the volume and not. Leave it off for music that has already been mastered, because it will override the mastering engineer’s decisions, and off for anything you intend to master yourself later.
AIFF, BWF and other uncompressed relatives
This page also accepts AIFF and AIFC (the Apple equivalent of WAV, common from Logic and older
Mac tooling) and Broadcast Wave files. BWF is a normal WAV with an extra bext
chunk holding timecode, originator and coding history — the audio decodes identically, though
the broadcast metadata has no MP3 equivalent and is dropped. If you need that metadata, keep
the BWF alongside the MP3.
A note on 32-bit float sources
Recorders like the Zoom F-series and MixPre write 32-bit float WAV, where samples can exceed 0 dBFS without clipping. Those convert here without trouble, but anything above full scale is limited when it is written to MP3. If your recording has genuinely hot peaks, pull the gain down in an editor first rather than letting the encoder decide where to clamp.
Related tools
To go the other way for editing, see MP3 to WAV. For a lossless archive at about half the size of WAV, convert to FLAC in the general audio converter. To stitch several session takes together with trims and crossfades before encoding, use the audio joiner.
Questions
My WAV is 900 MB. Will this handle it?
This is the case the tool is built for. Nothing is uploaded, so there is no 100 MB or 500 MB ceiling and no waiting in a queue — the only real limit is your device’s memory, and the browser sandbox tops out around 2 GB of working memory. On a phone, expect trouble past a few hundred megabytes.
How much smaller will it get?
Stereo 44.1 kHz 16-bit WAV is 1411 kbps. At 192 kbps MP3 that is about 14% of the original, and at 320 kbps about 23%. A one-hour interview drops from roughly 600 MB to 85 MB at 192 kbps. The result panel shows the exact ratio for your files.
What bitrate for a podcast or interview?
96–128 kbps mono is the podcast standard and sounds fine for speech, and forcing mono halves the size again if the recording is a single voice. Use 192 kbps or higher only when there is music, room ambience or multiple mics you want to keep separated in the stereo field.
Should I normalise the loudness?
Turn it on for interviews and voice recordings captured at inconsistent levels — the EBU R128 filter brings everything to −16 LUFS, which is what podcast platforms expect. Leave it off for music that has already been mastered, because it will undo the mastering engineer’s decisions.