We needed to turn a few photos into one PDF. We found a free online tool, and the first try worked. Then we wanted to rotate a page and make it again, and a pop-up was waiting: "You've reached your free limit. Sign up to continue."
So we signed up, just to finish one small job. We only need tools like that now and then, so giving in is always the quickest way out. And we hated it every time.
Now, with LLMs, building the tools we need ourselves is finally practical. So we built them, and we're sharing them the way we always wished we'd found them: free however often you use them, with no signup and no hidden limits.
Running everything in the browser is what makes that possible. Signup walls usually exist for a reason: the tool processes your file on its own server, where every use costs money, so it rations them. When the work happens on your device instead, your tenth use costs the site no more than your first: next to nothing. And your files never leave your computer or phone.
Most of these tools exist because we needed them, not for the sake of building something (well, except one or two). The job that started it all, photos into a PDF, is now a tab in our Image Toolkit: rotate, reorder and redo it as often as you like.
The tools
They all live on misc9.app. The first four are the ones we use most.
- š PDF Builder: merge PDFs, keep only the pages you want (
1, 3-5, 8) and add images as pages - š§ JSON Viewer & Diff: format, minify, sort keys and compare two files key by key
- š° Loan Calculator: work out your monthly installment, how much you can borrow and what paying early saves
- š¼ļø Image Toolkit: compress to a target size, convert iPhone HEIC photos, resize, make passport photos and turn images into a PDF
- āļø PDF Editor: add text and signatures, rewrite words already on the page, fill in forms, redact and reorder pages
- āļø Background Remover: cut the background out of a photo with AI and download a transparent PNG
- š” Wheel Spinner: spin to pick a random name, choice or chore
No account, no watermark, no upload. Every tool also works in 7 languages: English, Spanish, French, German, Dutch, Portuguese and Hindi.
Two versions of the same API response, with the keys in a different order on each side. Only the 7 real differences show up: new in green, missing in red and changed in amber.
Under the hood
- A static Astro site with React islands. Every page is plain HTML, and each tool is a single interactive island.
-
Heavy libraries load only when needed.
pdf-lib,pdfjs-dist,onnxruntime-weband the HEIC decoder come in through dynamicimport()the first time a tool needs them, so the first visit stays light. - The Background Remover runs AI in your browser. U²-Net (the 4.6 MB "p" version) runs in WebAssembly through ONNX Runtime Web. The model is fetched the first time you use it, and your photo is processed on your device.
- Redaction removes the text for real. A black box drawn over text usually leaves the words in the file, where anyone can select or search them. The PDF Editor turns a redacted page into an image with those areas painted out, so the original text is gone. The price: that page is no longer searchable.
- The JSON diff pairs by key, not by line. That's why the reordered keys in the screenshot don't count as changes. Broken JSON still opens too: trailing commas, single quotes, comments or a file cut off mid-write are recovered, and the first real error is shown with its reason.
- The honest trade-off: your device does the work. A very large file is only as fast as your machine.
Don't take our word for it
Open DevTools ā Network, drop a PDF or a photo into one of the tools and watch the list. You'll see the page load its own code (plus the AI model, the first time you use the Background Remover) and a cookie-free visit counter from Cloudflare. What you won't see is your file going anywhere.
Tell us what's missing
What's missing, and which of these would you actually use? And if some tool keeps putting a signup wall in front of you, tell us which one. It might be the next one we build.

Top comments (0)