OnWebChange vs changedetection.io for developers, plus a hosted AI option

Developers comparing OnWebChange vs changedetection.io usually want to know three things: how each one targets content, how alerts reach their own systems, and who has to keep it running. This comparison covers both tools on those points, then adds a hosted AI monitor for the cases where neither selector-based approach holds up.

The short version

Prices from each vendor's site, October 2026.

OnWebChange for developers

OnWebChange is the cheap, hands-off choice. You define a tracker in the web UI, point a CSS selector at the element you care about, and add rules, filters or number conversions. It renders JavaScript pages, can send custom headers to reach content behind a login, and lets you choose tracking locations in different regions.

For integration, every plan includes URL callbacks and a personal RSS feed of tracker content. On Standard and Turbo plans you also get CSV import and export, tracker groups and history with line graphs for numeric values. Checks per month come from one pool that you split across trackers, and the fastest checks run about every 5 minutes.

What you do not get is a documented API for creating trackers from code, at least not one listed on its pricing or features pages.

changedetection.io for developers

changedetection.io is the tool you run yourself. It is open source, installs with Docker or pip, and targets content with CSS or XPath. Browser steps let it log in or click through a page before reading it, and notifications go out through a long list of services. It has its own API.

The cost is your time. You host it, update it and fix it when a site changes its markup. The hosted plan at $8.99 a month removes the server work but not the selector work.

Where both run into the same wall

Both tools watch a location in the DOM. That is precise until the site ships a redesign, at which point the selector points at nothing or at the wrong block. Both also decide what counts as a change with text or keyword logic: a word appears, a number crosses a value, a line differs. If you have ever written an ignore rule for a timestamp and then another for a cookie banner, you know the pattern.

A hosted AI option

AyeWatch takes a different route. You give a webpage monitor one or more URLs and a sentence, for example "notify me when a deprecation notice or a breaking change appears in these release notes". AyeWatch judges each change against that sentence and sends a short summary when it matches. There is no selector to fix after a redesign.

For developers the useful parts are on paid plans:

• REST API to create, list, update and delete topics, which is what AyeWatch calls its monitors.
• Signed webhooks (HMAC) with topic.update events that include the summary.
• An MCP server, so an AI coding agent can set up a monitor from a prompt like "watch this changelog and tell me when a new version ships".

Paid plans start at $9 a month, after a Free Preview with no card. For a developer, the draw is that one plain-English rule replaces selector upkeep, and the signed webhooks carry a summary your code can use as is.

Which one to pick

• Stay on OnWebChange if you need cheap, stable CSS tracking, PDF or file tracking, or header-based access to login pages.
• Use changedetection.io if you want to own the stack, need browser steps, or must keep data on your own server.
• Use AyeWatch if your trackers break on redesigns or flood you with alerts, if you want to create monitors from code or an agent, or if you need to watch topics across news, forums and social sources as well as pages.

Many teams end up with two: a selector tool for the few exact values, and AyeWatch for everything where meaning matters more than position. The AyeWatch comparison page for OnWebChange lists the features side by side.