HTML checks and Browser checks.

Which loading method should I choose?

HTML checks: Read content delivered in the page HTML, without running JavaScript. Start here when the content you need is present in the preview.

Browser checks: Run JavaScript in a browser to capture dynamically loaded content. Try this when the expected value is missing from an HTML preview. Browser checks use a separate usage allowance and schedule.

Price thresholds, stock transitions, text conditions, and new-listing rules work with either method. The loading method does not determine the target type. Use the guided setup to confirm up to three targets from one page load. Selection and rule guide →

Choose a target

Whole page captures page text. Main content uses the first main, article, or role=main element. Specific element uses a CSS selector such as #price, .job-list, or [data-testid="stock"]. New guided selections must match exactly one element. Legacy text selectors still combine up to 100 matches. Standard CSS and bounded nth-of-type selectors are supported; other pseudo-classes and escapes are not.

A missing target is a check issue, never an empty “everything was deleted” notification. Review the current captured text on the monitor’s detail page. Changing the loading method or target creates a fresh baseline; old history stays within its retention and storage limits.

Browser checks by plan

Free includes a one-time 7-day browser trial: 1 monitor and up to 168 hourly scheduled attempts. It starts when you save your first browser monitor; previews do not start it. Deleting the monitor, changing settings, or crossing a monthly reset does not renew the trial. The monitor pauses when time or attempts run out, without switching your chosen loading method. Paid subscribers do not receive another free trial after cancellation.

Starter includes 750 Browser checks per UTC calendar month, at 60-minute intervals or slower. Pro includes 5,000 at 15-minute intervals or slower. HTML and browser monitors share your plan’s total slots; the check allowances are separate. Failed checks and retries count. Setup previews use their own allowance.

Service capacity, daily limits, website backoff, and the queue can delay checks. HTML checks continue independently. On a downgrade, the new allowance and minimum interval apply; checks wait if usage already exceeds the allowance. Losing browser access pauses those monitors. You can upgrade or explicitly switch them to HTML checks.

Which dynamic pages work?

Browser checks support public, read-only pages that load content through ordinary HTML, JavaScript, and GET requests. The browser waits for the target, then for its text to settle. It uses a consistent desktop viewport, English locale, and UTC timezone.

Sign-ins, supplied cookies, recorded clicks, form submissions, POST-based APIs, infinite scrolling, embedded frames, shadow DOM, WebSockets, service workers, downloads, paid proxies, and CAPTCHA solving are not supported. Scheduled checks skip images, video, and fonts. Visual setup previews may load images for context. Write requests, including analytics beacons, are blocked. If a page needs a POST request for its content, that content will not load.

Some websites deny automated access. Delayed or continually changing text may require a narrower selection or may be unsupported. Confirm the value in your captured preview before saving. The visual picker selects page content; PageFluxer does not compare screenshot pixels.

Page download and capture limits

HTML text previews and scheduled HTML checks can download up to 5 MiB of HTML. Scripts, styles, comments, and unsupported embedded content are removed before the cleaned HTML is checked against a 1 MiB capture limit. Selected content and retained history keep their existing limits. Pages are never truncated to make a capture fit.

Visual previews and Browser checks use the renderer’s separate request and transfer limits. Switching to text preview can help when preview images make a page too heavy. A specific element narrows what is monitored, but the page still needs to load first.

Understand a check issue

• TARGET_MISSING: the selected content could not be found. Review the selector.
• PAGE_NOT_READY: the target text did not settle before capture.
• PAGE_BLOCKED: the page returned an error or an access challenge.
• NETWORK_BLOCKED: a request was disallowed or failed.
• RESOURCE_LIMIT: the page exceeded request, transfer, or content limits.
• RENDER_TIMEOUT: the browser could not finish in time.

A failed check preserves your last successful capture. Check the URL and selection before retrying; repeated failures back off. If the service is unavailable, checks wait until it is restored.