A WAV file is uncompressed PCM, so there is no smaller WAV to be had by re-saving it unchanged. Compressing a WAV means encoding it into a lossy format such as MP3, M4A or OGG, which cuts the size by a large margin and permanently discards audio detail. Keep the original WAV.
The size of a WAV file is set by its sample rate, bit depth and channel count, not by a quality slider, so nothing inside the file can be thrown away without decoding it first. Every option in a browser compressor therefore ends in MP3, M4A or OGG. A 44.1 kHz, 16-bit stereo WAV runs at roughly 10 MB per minute; the same minute at 128 kbps lands near 1 MB. The sections below cover which setting to choose, when a target size is reachable, and the jobs where you should keep the WAV untouched.
People search for a WAV compressor expecting the file to stay a WAV and get smaller. No tool can do that by re-encoding, because WAV stores raw samples with no compression to remove. What you can do is create a smaller copy in a delivery format and keep the original for editing.
Audio Compressor
Add a WAV, pick a bitrate or a target size, and download MP3, M4A or OGG. The encoder runs in the tab, so the recording is never uploaded.
Why a WAV file is so large
WAV is a container for PCM audio, which means every sample is written at full size. A stereo recording at 44.1 kHz and 16 bits holds 44,100 samples per second per channel, and each sample takes two bytes. That comes to about 176 KB per second, or roughly 10 MB per minute, before any file system overhead.
None of those bytes is redundant in a way a general compressor can remove. Zipping a WAV saves almost nothing because the samples are close to random from an archiver point of view. Lossless codecs such as FLAC do better by predicting the waveform, but they still keep every sample, so the saving is modest next to a lossy encoder.
The consequence is that size grows in a straight line with duration. A three minute track is about 30 MB, a one hour interview is about 600 MB, and a two hour recording passes a gigabyte. The table below is arithmetic from the sample rate and bit depth, not a measurement of a specific file, and it changes with the channel count.
| Duration | WAV, 44.1 kHz 16-bit stereo | MP3, 128 kbps |
|---|---|---|
| 1 minute | about 10 MB | about 0.9 MB |
| 3 minutes | about 30 MB | about 2.8 MB |
| 10 minutes | about 101 MB | about 9.4 MB |
| 60 minutes | about 606 MB | about 56 MB |
The WAV column is fixed by the recording settings, and the MP3 column by the bitrate.
Two ways to make it smaller: a bitrate, or a target size
The compressor accepts MP3, WAV, M4A, AAC, OGG, FLAC and Opus as input, and offers MP3, M4A and OGG as output. A WAV goes in and a lossy file comes out. There is no WAV option in the output row, and no FLAC either. Once that is clear, there are two controls.
Target bitrate
A fixed list of 64, 96, 128, 192, 256 and 320 kbps. The lower the number, the smaller the file and the more detail the encoder drops. 128 kbps is selected by default, and it is the usual balance for spoken word.
Target file size
You enter a number in megabytes, from 0.1 upward, and the tool divides the bits that fit by the duration. The result is clamped between 32 and 320 kbps, and the line under the field reports the bitrate it will actually use, including a note when the target falls outside that range. The estimate line reads "Estimated output: about 2.1 MB at 128 kbps MP3".
That estimate is duration multiplied by bitrate, so headers and encoder padding can shift the real file slightly. It is close enough to decide whether to run the job, not a promise of the exact byte count.
| Mode | What you choose | What the tool decides | Best for |
|---|---|---|---|
| Target bitrate | One of six values | Nothing, it uses your value | Predictable quality across files |
| Target file size | A size in MB | The bitrate, clamped to 32 to 320 kbps | Upload caps and email limits |
How to compress a WAV file in three steps
- Add the WAV. Drop it on the tool or click to browse. The duration is read on your device, and a target size cannot be calculated until it is known.
- Choose a mode and an output format. Leave the mode on Target bitrate and pick 128 or 192 kbps, or switch to Target file size and enter the limit you have. Pick MP3 if the file has to play anywhere.
- Compress, listen, download. The progress bar shows the encoder working, then a player appears with the result and its size. Listen before you keep it, because this is where you find out whether the bitrate was too low.
The first run downloads the FFmpeg engine, about 30 MB, which the browser caches afterwards. That is the engine, not your audio. One file is processed at a time, so a folder of WAVs has to be added one by one.
Which bitrate to pick
The tool page recommends 128 kbps for speech and 192 kbps or higher for music. That matches what most listeners notice on ordinary earbuds.
| Bitrate | Sensible for |
|---|---|
| 64 kbps | Voice memos where words matter more than tone |
| 96 kbps | Lectures, interviews, long speech |
| 128 kbps | Default. Speech and casual listening |
| 192 kbps | Music, the usual first step up |
| 256 kbps | Music that will be played on good equipment |
| 320 kbps | The most detail this tool can keep |
Two cautions. Bitrate sets how much data the encoder may spend per second, and it cannot restore detail the source never had, so a 320 kbps export of a noisy phone recording will not sound like a studio master. And a lower bitrate is not automatically a failure: 96 kbps is usually fine for an interview through earbuds, while the same setting on a string quartet is obvious. The full guide to audio compression goes deeper into how the number maps to file size.
What you actually lose
- The WAV container. There is no WAV output, so the file you download is a different format with a different extension.
- Audio detail. All three outputs are lossy, and the loss is permanent in that copy. Re-compressing an already compressed file adds another generation of loss.
- Control over sample rate, bit depth and channels. There are no fields for them, so you cannot deliberately downmix a stereo WAV to mono or lower the sample rate to halve the size.
- Batch processing. One file per run, with no queue.
- A guaranteed size. The target mode clamps the bitrate, so a number outside the range is replaced by the nearest usable value.
- A way back. The original WAV stays untouched, but the compressed copy cannot be turned back into what it was.
Compressing a WAV to 2 MB, and when the number is reachable
A 2 MB target is a duration problem more than a setting problem. Two megabytes is 2,097,152 bytes, or 16,777,216 bits. At the 320 kbps ceiling that lasts about 52 seconds; at the 32 kbps floor it lasts about 8 minutes 44 seconds. Outside that window no combination of settings will produce a 2 MB file, because the tool will not go below 32 kbps or above 320 kbps. This is arithmetic, not a measurement, and real files carry a little overhead.
| Bitrate used | Longest audio that fits in 2 MB |
|---|---|
| 320 kbps (maximum) | about 52 seconds |
| 128 kbps | about 2 minutes 11 seconds |
| 64 kbps | about 4 minutes 22 seconds |
| 32 kbps (minimum) | about 8 minutes 44 seconds |
If your recording is longer, the honest options are to trim it or to accept a file above 2 MB. The Audio Trimmer cuts the part you need before compression, and this walkthrough covers the order of operations. Compressing first and trimming second wastes a generation of encoding on audio you were going to delete.
One more detail: the estimate line updates as you change the size field, so you can watch the clamp happen. Type 0.2 MB for a one hour interview and the field will report that it is using 32 kbps, while the estimate stays far above 0.2 MB. That is the tool telling you the truth rather than failing silently.
When you should not compress a WAV
There are jobs where the WAV is the point and a lossy copy is the wrong deliverable.
- Editing and mixing. Keep the WAV through every edit and export a lossy copy at the end. Compressing before editing means every later export starts from damaged audio.
- Masters and archives. A master is the file you keep to make future versions from. Store the WAV, or a FLAC copy if disk space matters, and compress only the distribution copy.
- Deliverables that must be WAV. Some broadcast, telephony and archival workflows specify PCM. This tool cannot produce that, so trimming duration is the only size lever it gives you. A lower sample rate and bit depth set at the recorder or editor also shrink a WAV, and that is outside this tool.
- Loudness or measurement work. Lossy encoding changes peaks and can shift what a meter reads.
If you need a lossless container, the Audio Converter can write FLAC, which stores the same samples in a compressed file. FLAC is usually smaller than the WAV it came from, though the saving depends on the material, and it plays in fewer places than MP3. That trade is right for an archive and wrong for a chat attachment.
Audio Converter
Write a WAV to FLAC for a smaller lossless copy, or to MP3, OGG and M4A for delivery. Runs in the browser, nothing uploaded.
Privacy: what stays on your device
The encoder is FFmpeg compiled to WebAssembly and it runs inside the tab. The WAV is read from local storage into page memory, decoded, re-encoded and handed back as a download. No request carries the audio, and no account is involved.
That matters for meeting recordings, client calls and unreleased audio, which are common reasons to have a WAV in the first place. You can check the claim by opening the browser developer tools, switching to the network panel and watching a job run: the only large transfer is the engine on first use. Our notes on browser audio conversion cover the same setup from the format side.