About
45 audio tools, no upload, no account, no ads.
Why we bothered
Every free audio tool on the web works the same way. You hand your file to a server, wait through an upload, wait in a queue, wait through a download, and dodge four advertisements on the way. Somewhere in there you hit a size cap, or a watermark, or a prompt to make an account.
None of that is necessary any more. Browsers have had a complete audio engine built in for over a decade. The same one behind web games and video calls. It can decode almost every format, process audio faster than real time, and write files back out. The server was never doing anything your own machine could not.
So these tools do the work locally. That removes the upload, the queue, the size cap and the privacy problem in one move, and it makes the tools faster than the paid alternatives.
How it actually works
Every page is static HTML with no framework runtime. The audio engine is tiered so that common tasks stay small: trimming, volume, fades, speed and effects use the browser's own audio nodes and download nothing extra. MP3 output loads a 110 KB encoder. Only genuinely non-native formats. M4A, OGG, FLAC, WMA, AIFF. Load a full converter, and only when you ask for one.
The practical result is that a fully working audio editor page weighs about 50 KB. Most comparable sites ship two to four megabytes before you have done anything.
What it's free of
No price, no trial clock, no "pro" tier holding the useful settings hostage, no watermark stamped on your file, no daily quota, no account, no email capture. You will not be interrupted by an ad, and you will not be asked to sign in to download the thing you just made.
That is not a launch promise with an expiry date. There is no premium version planned, and if anything ever changes, every tool here keeps working exactly as it does today.
The unreasonable parts
Some of this took much longer than it looks. The waveform draws one reading per physical pixel and layers a peak outline over an RMS body, so a single stray click cannot make a quiet passage look like a chorus. The loudness meter implements the actual broadcast standard, K-weighting and both gating stages included, so the number it gives you matches what Spotify measures rather than being roughly in the area.
The tempo tools use real time-stretching instead of just resampling, which is the difference between slowing a song down and making everyone sound like they are underwater. Nobody is going to notice any of this individually. That is sort of the point: the work is there so you never have to think about it.
What's not here yet
Some things genuinely need more than a browser can do well today. AI-based vocal removal and stem separation need large models; automatic transcription needs a speech model measured in hundreds of megabytes. Those are on the list, and they will only ship if they can run locally like everything else.
Multi-track editing is deliberately absent. These are single-purpose tools: one job each, no learning curve. If you need a timeline with layers, you need a DAW, and there are good free ones.
Tell us what to build
If a tool is missing, or one of them mangled your file, that is worth knowing about. The tools that get built next are the ones people actually ask for.