Own your music again.

The universal standard. Decoded by our bespoke Fonix MP3 engine — a clean-room implementation purpose-built for the ESP32, with hand-tuned Xtensa assembly in the hottest loops. Supports all MPEG-1/2/2.5 Layer III sample rates with near bit-perfect output.

Lossless audio, bit-for-bit identical to the original recording. Fonix One uses a finely tuned custom version of the miniflac decoder, delivering full-quality playback with the smallest memory footprint of any format we support.

The modern successor to MP3, offering better sound quality at lower bitrates. Our bespoke Fonix AAC decoder handles AAC-LC at all standard sample rates, built from scratch with a custom radix-4 FFT core and hand-tuned Xtensa assembly.

AAC audio wrapped in an Apple-style container. Our custom-built M4A demuxer streams the file structure on the fly rather than loading it all into memory, keeping things fast even on longer tracks.

Video container, audio soul. Fonix One extracts and plays the audio track from MP4 files using the same smart streaming demuxer as M4A, handy for podcasts and downloaded media.

Raw, uncompressed audio with zero processing overhead. Our minimal WAV parser reads the PCM data straight from the file with practically no memory cost, as close to the source as it gets.

The modern open codec behind YouTube, WhatsApp, Discord, and a growing share of podcasts. Our bespoke Fonix Opus decoder implements all three modes — SILK for voice, CELT for music, and hybrid — entirely in fixed-point, with output that matches the reference decoder sample-for-sample.