We built the audio editor that can't leak your files.

The problem we kept running into

Almost every "online audio editor" works the same way under the hood: you hand over your file, it travels to a server you know nothing about, it gets processed there, and a result comes back. For a meme clip, fine. For a voice memo, an unreleased track, a recorded interview, a confidential client deliverable, or a family recording you would never want leaked — that quiet upload is a real cost most people never agreed to and rarely even notice.

The frustrating part is that the upload is usually unnecessary. Modern browsers are extraordinarily capable. The same math that a server runs to trim, convert, or compress audio can run just as well on the device already sitting in front of you. The upload step persists out of habit and business model, not technical necessity.

So we built the opposite

AudioDock does all of its work in your browser. When you open a tool and drop in a file, the processing happens on your machine using the Web Audio API. For the conversions that genuinely need a codec — encoding an MP3, writing an OGG, pulling audio out of a video — we use ffmpeg compiled to WebAssembly, which runs that same battle-tested engine locally instead of on a server.

The result is a toolkit that feels like a web app but behaves like desktop software: no accounts, no upload bars, no queue, no file-size ceiling imposed by someone else's infrastructure. A large recording is processed as fast as your computer can do the math, and a sensitive file never touches hardware you do not control.

Privacy you can verify, not just trust

Plenty of products promise to respect your privacy. We would rather you not have to take our word for it. Because AudioDock never uploads your audio, you can prove it: open your browser's developer tools, watch the Network tab, and run any tool. Your file does not appear, because it never leaves the tab.

That is the difference between privacy as a policy and privacy as an architecture. A policy is a promise that can change. An architecture is a fact you can check. We chose the second one on purpose.

What we believe

Good tools should be fast, honest about their limits, and respectful of the person using them. We would rather ship a smaller set of tools that genuinely work than a long menu of half-finished features. Where a tool depends on something heavier — like the one-time download of the ffmpeg module — we say so plainly instead of hiding it.

We also believe the default should be your control. Nothing about your files is stored after you close the tab, optional conveniences like a local activity log are off until you turn them on, and a single click wipes everything AudioDock keeps on your device.

Where it goes from here

AudioDock is built to grow without compromising the core idea. Loudness normalization, noise reduction, and a richer waveform view are on the roadmap, and every one of them is being designed to run locally, the same as everything else. If a future feature ever genuinely required a server, we would treat that as a serious decision and tell you exactly what it meant — not slip it in.