About

What this site is

anaitool is 30 free tools for developers, people building on language models, and anyone doing technical SEO. Each one does a single job: convert between two data formats, count tokens, generate markup, inspect a file. There is no account, no usage limit and no paid tier holding back the useful half.

It is run by Kyle Maddie, working from Karachi. It started because the existing options for these tasks were mostly the same: an ad-heavy page wrapping a thin script, asking for an email address before showing a result, and quietly uploading whatever you pasted to a server. That seemed like a solvable problem.

Why the tools run in your browser

Almost every tool here processes your data entirely in the page, using JavaScript. Nothing is transmitted. You can verify this yourself: open your browser's network panel, use any tool, and watch for a request that never comes.

This is not a marketing line, it is the main design constraint, and it rules things out. Some features would be easier with a server — better file handling, heavier processing, saved history — and they are not here because of it. The reason is that the data people paste into tools like these is frequently sensitive: Kubernetes secrets into a YAML converter, production JWTs into a token decoder, a client's customer list into a CSV converter, an unpublished manuscript into a word counter. Most tools that accept that data send it somewhere, and most people never think to check.

Where a tool genuinely cannot work client-side, its page says so plainly rather than leaving you to guess.

How it is funded

Honestly: it currently is not, beyond covering hosting. There are no ads on the site today. If that changes, it will be stated here, and it will not involve tracking scripts that follow you around the web or anything that transmits what you paste into a tool.

The site does not use analytics that identify you. There is a first-party pageview counter that records which pages are visited, roughly how many people visit, and where they arrived from — no cookies, no cross-site identifiers, and no way to single anyone out. The privacy page describes exactly what is recorded.

How the tools are built

Plain PHP on the server, plain HTML, CSS and JavaScript in the browser. No framework, no build step, no npm dependency tree. Pages are small and load fast because there is very little to load.

Each tool is a self-contained folder with its own interface, logic and written explanation. That structure exists so a new tool can be added without touching anything else, which in turn means tools get added when they are useful rather than when there is time for a release.

Where the explanations come from

Every tool page carries a written explanation of the format or problem it deals with — not padding for search engines, but the things that actually catch people out. YAML silently turning Norway's country code into false. Base64 being mistaken for encryption. Structured data with invented review ratings earning a manual action. Those sections exist because getting them wrong is common and the consequences are real.

Where something is uncertain or contested, the page says so. Where a widely repeated claim is wrong — that there is an optimal keyword density, that llms.txt improves rankings — the page says that too, even when the tool on that page is the one people came for.

Corrections and suggestions

If something on this site is wrong, please say so — contact page. Technical corrections are especially welcome and get fixed quickly. Tool suggestions are read, and the ones that come up more than once tend to get built.