Audio file size is almost entirely a function of bitrate — how many kilobits of data represent each second of sound. A CD-quality WAV file stores roughly 1,411kbps uncompressed; an MP3 at 320kbps is already a quarter of that with no audible difference for most listeners, and at 128kbps it's a ninth of the original size while still sounding reasonably clean on earbuds or a phone speaker. This tool re-encodes any audio file through libmp3lame at one of three fixed bitrates, trading some of that technically-measurable fidelity for a dramatically smaller file.
The three presets map to real use cases rather than abstract quality labels. High Quality (192kbps CBR) is close to what most streaming services use for their default tier — music retains its dynamics and detail, and the size reduction mostly comes from discarding frequencies and stereo information the ear doesn't rely on much. Medium Quality (128kbps) is the long-time standard for 'good enough' MP3s — podcasts, audiobooks, and spoken-word content lose essentially nothing perceptible at this rate, and even most music sounds acceptable on consumer playback gear. Low Quality (64kbps) is aimed squarely at speech — voice memos, lecture recordings, meeting audio — where intelligibility is all that matters and the bitrate most listeners would reserve for music becomes unnecessary overhead.
Unlike the variable-bitrate quality settings used by format-conversion tools, this compressor uses constant bitrate (CBR) encoding on purpose: CBR produces a predictable, calculable output size (duration × bitrate), which matters when you're compressing specifically to fit under an email attachment limit or a platform's upload cap rather than just wanting smaller files in general. Every preset outputs a standard MP3, the most universally playable audio format, regardless of what format you uploaded.
Need a different output format instead of smaller MP3? format conversion tools change the container; this tool specifically targets file size.
