QiLing ToolsWhy Your Files Never Get Uploaded - How In-Browser Processing Works
QiLing Tools' PDF, image, audio and video tools run inside your browser, so files never touch a server. Here is how that works, how to verify it yourself, and its limits.
Most online tools work like this: you upload a file to their server, the server processes it, and you download the result. QiLing Tools works the other way around — the processing code is downloaded to your browser and runs on your own device, so your file never leaves it.
How a browser can process PDFs and videos
Browsers used to be good only at showing web pages; heavy file processing had to happen on a server. That has changed. A few technologies now give the browser close to desktop-app capabilities:
- WebAssembly compiles mature C/C++ programs into a format the browser runs at near-native speed. That is how FFmpeg (for audio and video) and libarchive (for RAR and 7Z) run on this site.
- JavaScript file libraries: pdf-lib rewrites PDF structure directly for merging, splitting and watermarking, and Mozilla's PDF.js renders pages for previews and PDF-to-image conversion.
- Canvas handles image compression, cropping, rotation and watermarks pixel by pixel, then encodes the result as JPG, PNG or WebP.
- On-device AI models: the segmentation model behind background removal runs through onnxruntime-web in your browser, using the GPU via WebGPU when available.
When you pick a file, the browser simply reads it into memory and hands it to these programs. The result goes straight to your downloads folder.
Verify it yourself
You don't have to take our word for it:
- Open any tool, for example Merge PDF.
- Press F12 (Option + Command + I on a Mac) to open developer tools and switch to the Network panel.
- Choose your files, start processing, and watch the new requests.
You will see requests that load the program itself (scripts and wasm files), but none whose size matches your file, and no upload request carrying its contents.
Why local processing is better
- Privacy: contracts, IDs and medical records never have to be handed to a third party, so there is no copy left on anyone's server.
- No queues or quotas: speed depends on your own device, not on server load, and we have no bandwidth costs that would force file-size or usage limits.
- Free with no sign-up: with no server-side processing costs, the tools can stay free and watermark-free.
The trade-offs
Local processing isn't free of costs. A few things are worth knowing up front:
- First use downloads an engine: audio and video tools fetch the ~30 MB FFmpeg core the first time, and background removal fetches a ~46 MB model. Your browser caches them afterwards.
- Device performance matters: video transcoding and AI background removal are noticeably slower on phones and older computers.
- Memory is finite: very large files (multi-gigabyte videos, for example) can exceed what the browser is allowed to allocate. Splitting the file first usually helps.
Summary
"Your files are never uploaded" isn't a slogan — it follows from how the tools run: the program comes to you instead of your file going to us. If any tool behaves differently from what this post describes, please tell us via the feedback button in the bottom-right corner.