I've been building free browser tools with one rule: your files never leave your device

A while back I was editing a video for my channel. I had filmed some footage with music playing in the background; nothing intentional, just the room. I went to upload it and immediately thought: copyright strike. I needed to strip the audio completely before dropping my own track on top.

I went looking for a quick browser tool to do it. Every option I found did one of three things: asked me to create an account, added a watermark to the output, or uploaded my file to a server I had never heard of.

I closed all the tabs and opened a terminal instead.

FFmpeg can remove audio from a video in one command. That is not the problem. The problem is that most people don't have FFmpeg installed, and most people are not going to install it just to mute one clip. So the market fills with these upload-based tools, because running software on a server is the obvious default when you want to build a web product.

But there is a thing called WebAssembly that has been sitting in browsers for years, and FFmpeg has been compiled to run in it. Meaning you can run FFmpeg inside someone's browser tab, on their own device, without sending anything anywhere.

I spent a weekend building that. The result is remove-audio.com. Drop a video in, hit remove, download the muted file. Everything runs locally. The files don't go anywhere. No watermark, no account, no paywall.

I made it completely free and put it online.

What I didn't expect

I assumed it would get maybe a few hundred visits from people in my network and then sit idle. That is what usually happens with side projects.

Instead it started picking up search traffic on its own. People searching "remove audio from video without uploading" or "mute video online free no watermark" were landing on it. These are people who already know what they want and are frustrated that everything requires a sign-up. The tool answers exactly that search intent.

The privacy angle matters more than I initially thought. I get messages from people who had confidential meeting recordings they couldn't upload anywhere. Editors dealing with client footage. Teachers who can't pass student videos through third-party services. The no-upload constraint I treated as a technical implementation detail turned out to be the main reason people chose it over the alternatives.

The pattern

After that I built the same thing for background removal: free-background-remover.com. Same principle. Everything runs in the browser, nothing goes to a server, completely free.

I now have a small collection of tools built around one constraint: your files stay on your device.

Why free?

The honest answer is that I don't need them to make money right now. I'm a software engineer, these are side projects, and I'd rather have a tool that thousands of people genuinely use and trust than one that makes $50 a month and annoys people with upgrade prompts.

There is also a selfish reason. Building browser-native tools that use WebAssembly is technically interesting to me. FFmpeg.wasm opens up a lot of things people assume require a backend. I am still exploring that space.

Where this goes next

The question I keep coming back to is: what happens when you add AI to this same model?

Most AI-powered tools send your content to an API. That makes sense for heavy inference. But a lot of the more common AI tasks, things like background removal, noise suppression, object detection, image cleanup, can now run inside a browser through smaller models and WebGPU. The gap between "AI feature" and "local AI feature" is closing.

I want to see how far that constraint can stretch. Can you build something genuinely AI-powered that still never touches a server? That is the direction I am headed with these tools: not smarter SaaS, but smarter browser software. Niched, local, free.

I do not have the answer yet. But the question feels worth chasing.

What I have actually learned

If you are going to build free tools, the thing that makes them defensible is not the features. It is the constraint. "No uploads, no sign-up, no watermarks" is legible to users, and legibility creates trust.

The number of people who say "I specifically chose this because I didn't want to upload my file" is genuinely surprising.

The "does this actually need a server?" question is now my default when I think about building any web tool.

More often than I expected, the answer is no.

— Robert Vaida