Why I stopped uploading files to random websites (and what I use instead)

I found https://filagram.com/ and this is what happened -

Last month I needed to cut the background out of a photo for a client deck. Opened the first Google result, dragged the image in, watched it fly off to a server in... wherever. Got my result. Looked fine. Moved on.

Then it hit me. That photo had someone's face in it. It just went to a server I know nothing about. Stored how long? Who has access? What are they doing with it?

This is what most of us do every week. Compress a PNG? Upload. Merge two PDFs? Upload. Convert HEIC to JPG? Upload.

We hand files to strangers constantly and never think twice.

The upload problem

Tools most developers have bookmarked:

• TinyPNG for image compression
• ILovePDF for PDF stuff
• CloudConvert for format conversion
• remove.bg for background removal
• Whatever random converter Google spits out first

All of them upload your file to a remote server. Some tell you about it. Some don't. Most privacy policies say something like "we may retain your file temporarily" which could mean anything.

For memes? Whatever. For client contracts, medical records, tax documents? Different story.

Browsers can handle this now

WebAssembly changed the math here. You can run the same compression algorithms that power server-side tools directly in a browser tab. Near-native speed, zero network round trip.

I went looking for tools that process files locally. Found some good single-purpose ones - Squoosh handles image compression well, pdf-lib works for basic PDF operations. But I wanted one tool that replaced the whole bookmarks folder.

Filagram does that. 70 tools covering images, PDFs, format conversion, background removal, and some developer utilities. All browser-based, nothing uploads anywhere.

The ones I reach for most:

• Image compression - drop a PNG in, get a smaller one back. Took a 4MB file down to 400KB without visible quality loss.
• PDF merge - combine documents without uploading them. Handles big files fine.
• Background removal - runs a full AI model locally in the browser. Takes a few seconds to load the first time, then it's fast. This was the one that surprised me.
• Format conversion - HEIC to JPG, PNG to WebP, the usual stuff.
• Dev tools - base64, JSON formatting, hash generators. Nothing fancy, just saves time.

Check for yourself

Open DevTools (F12), go to the Network tab, clear it, then process a file. If you see your file getting sent as a POST request somewhere, that tool uploads your files. If the network log stays empty during processing, it's running locally.

I ran this test on Filagram. Nothing goes out during processing. The only network request is the AI model file for background removal, which downloads once and runs locally after that.

It's not perfect

Big files (50MB+) can lag on older hardware. Server-side tools have more raw compute available. And some things like OCR still work better in the cloud.

But most daily file work - compressing, converting, merging - happens faster locally because there's no upload/download overhead. And your files stay yours.

Who actually needs to care about this?

If you handle client data, probably you. Healthcare, finance, legal - definitely. If you're compressing screenshots for a blog post, probably doesn't matter.

But once you start paying attention to where your files go, it's hard to stop. I check the network tab reflexively now.

Filagram has a free tier (3 uses/day). $5/mo for unlimited if you need it. But the bigger point is worth thinking about regardless of which tool you pick. Open your bookmarks and check - how many of those file tools are uploading your stuff to servers you've never looked into?