A changedetection.io alternative built for systems
changedetection.io is an excellent self-hosted page monitor for a person watching a handful of pages. Hypeline solves the neighbouring problem: turning many sources into a clean event stream that your own software consumes, at scale, over an API.
Different jobs, honestly compared
| changedetection.io | Hypeline | |
|---|---|---|
| Built for | A person watching pages, via a dashboard | Systems consuming events, via an API |
| What triggers an alert | Any change in the watched selector or text | Genuinely new content, after extracting the main text |
| Output | Notifications (email, chat, and more) | A normalized event stream: SSE, cursor catch-up, signed webhooks |
| Duplicates | Each check can re-alert on the same change | One event per real change, never re-sent |
| Beyond web pages | Web pages (plus RSS watching) | Pages, every feed format, and live push sources in one stream |
| Scale model | You host and tune your own instance | Thousands of sources with adaptive scheduling, managed |
| Event authenticity | Not applicable | Ed25519-signed events you verify with a public key |
Fair is fair: if you want a self-hosted monitor with visual diffs for a personal set of pages and a human reading the alerts, changedetection.io is the right tool. Reach for Hypeline when the reader is your code.
Frequently asked questions
Is Hypeline a drop-in replacement for changedetection.io?
Not exactly, and it does not try to be. changedetection.io is a monitoring tool a person configures and reads. Hypeline is an API: it turns watched sources into a deduplicated stream of events that your own software consumes over SSE or signed webhooks.
When is changedetection.io the better choice?
When you want a free, self-hosted monitor for a personal or small set of pages, care about visual and selector-level diffs, or want alerts in front of a human without writing any code. It is excellent at that job.
When is Hypeline the better choice?
When the consumer is software rather than a person: you need clean deduplicated events at scale, one schema across pages, feeds, and push sources, cursor-based recovery after downtime, and delivery into pipelines, queues, or agents.
How does Hypeline avoid the noise page monitors alert on?
It extracts the main content from each page before comparing fingerprints over time, so ads, template edits, and timestamp churn do not become events. Only substantive new content does.
Be first to make the static web live
Add your email and we'll tell you the moment it goes live.
No spam. One email when it goes live.