Know when your public API response changes.

A service may continue returning HTTP 200 while a published status changes, a configuration value disappears or a reference list is updated. RegressionRadar's HTTP content probe records the status and hashes the response text, so either kind of change can begin a new streak.

API monitoring uses the same HTTP content method as webpage monitoring. It compares the returned text; it does not validate a JSON schema or assert that a field has a particular value.

1. 01

   Choose the response to watch

   Use a publicly reachable endpoint URL, such as a service-status JSON feed. Avoid responses containing a timestamp, request ID or random value on every request.

2. 02

   Create an HTTP content monitor

   Enter a name and URL, choose HTTP content, and leave the HTML selector blank for JSON. Set the interval and timeout, then enable the monitor.

3. 03

   Inspect the baseline

   Review the first measurement's status and available stored body. The first-ever streak sends no change notification. Confirm that you are getting the intended payload rather than an access-denied page.

4. 04

   Follow status and body changes

   Matching status and content hash continue a streak. Review new streaks, inspect measurement failures and compare stored bodies with the text diff. Add email or Teams notification rules for the changes you want to follow.

A public service-status feed

Before: { "service": "checkout", "status": "operational" }
After:  { "service": "checkout", "status": "degraded" }

Both illustrative responses can return HTTP 200. The body change still starts a new streak. A transition to HTTP 503 also changes the status and is recorded as an unsuccessful check.

A published configuration or reference list

Track a public response listing supported regions or a published version. A changed list or version alters the response hash. Review the text diff to decide whether consumers need an update.

A predictable probe target

Use an endpoint that returns only a period-stable value when trying the workflow. RegressionRadar's hour-of-day test endpoint returns the same payload within an hour. It is a demonstration target, not a service-health assertion.

• The monitor sends GET requests. The setup does not offer POST bodies, custom authorization headers, OAuth flows or scripted request sequences. Choose endpoints reachable without those requirements.
• JSON is compared as response text. Whitespace, property order and changing timestamps can trigger new streaks even when the meaningful data is equivalent.
• There are no JSONPath selectors, schema assertions or per-field ignore settings. HTML CSS selectors require text/html and cannot filter JSON.
• Response time is recorded, but a latency threshold is not part of the status-and-content streak comparison. A timeout produces a failed measurement.
• Responses over the configured download limit fail. A body may be hashed without being stored if it exceeds the storage limit; text comparisons require available stored content.
• Scheduled checks can miss an outage or change between runs. Alerts depend on configured streak rules and available credits.

Each HTTP check uses 1 credit, including failed checks. Each successfully sent email or Teams notification uses 1 credit. Failed or skipped notifications use none.

For one URL over a 30-day month, daily checks use 30 credits; hourly checks use 720 credits, before alerts. Actual totals depend on the number of days and completed checks. Adding monitors multiplies the check budget.

The Free plan includes 50 credits per month: up to 50 checks if you spend the entire allowance on this method and send no notifications.

• Bronze: 1,000 credits per month at US $9.99 per month, excluding VAT.
• Silver: 10,000 credits per month at US $19.99 per month, excluding VAT.
• Gold: 100,000 credits per month at US $29.99 per month, excluding VAT.

Credits are shared across all monitors and sent notifications and reset each calendar month (UTC). An action needs enough credits for its full cost; when the allowance is exhausted, checks and alerts pause until the next month or a plan upgrade. A paid plan has an end date; Free limits apply after expiry.

Compare plans, trial and extension terms and review usage forecasts in your account before increasing check frequency.