Continuous reachability monitoring of the major sports odds APIs, ours included. Run by ParlayAPI, one of the providers measured, with the methodology next to every number. Machine-readable at /watch.json.
How to read this page: Reachability of public endpoints measured from our production server in Falkenstein, Germany, every 5 minutes. This is not an official status page for any provider listed, ours included. A 401 from a keyed endpoint counts as reachable: the vendor's edge answered, we just did not present an API key. Our own API is measured the same way and shown the same way.
A target counts as down only when it returns no HTTP response within 10 seconds or answers with a 5xx.
One request per endpoint per run, from one vantage point: these numbers are our probes' view of public-endpoint reachability, not any provider's true global availability.
For the authoritative word on any provider, check that provider's own status page.
API edge. Answers 401 without a key; a 401 counts as reachable.
24h reachability, our probes100.0%of 288 probes, last 24h
last probe216 ms, HTTP 4013 min ago
latency of reachable probes, last 24h: min 152 ms, max 8,375 ms
Data generated 2026-09-10T03:23:03.839568+00:00. Cached for 60 seconds; probes run every 5 minutes.
Methodology
Vantage point: ParlayAPI production server, Falkenstein, Germany.
Cadence: one GET per endpoint every 5 minutes, 10 second timeout, identifying itself as ParlayAPI-watch.
Up: any HTTP response with status under 500 within the timeout; 401/403 from keyed endpoints count as reachable.
Down: no HTTP response within the timeout (timeout, DNS, connect, TLS failure) or a 5xx response.
State: last 3 probes: all up = up, none up = down, mixed = degraded.
Endpoint selection follows the public odds-api-benchmarks probe configuration: each vendor's most public API-edge endpoint.
one vantage point measuring public-endpoint reachability; not a provider's true global availability and not a statement about data quality or odds update speed.
latency is measured from our own production network, and the ParlayAPI endpoints run on the same server that sends the probes, so our latency numbers have a home advantage; latency is not comparable across providers on this page.
Who runs this: ParlayAPI, which is one of the providers measured on this page; this is not third-party monitoring.
This page exists because "is my odds API down or is it just me" is a real question at kickoff time, for users of every provider here. If a provider believes their endpoint is misrepresented, email [email protected] and we will fix the probe. The same address gets an odds API added to this page. Provider names and endpoints belong to their respective owners; this page is not affiliated with, or endorsed by, any provider listed.