Is OpenRouter Down?

No — OpenRouter is up

Reachable from all 8 checked regions

Average response time: 57ms

Last checked 1 hour ago · checks run every 6 hours

Official status page: https://status.openrouter.ai

OpenRouter uptime

100%

Last 7 days

100%

Last 30 days

100%

Last 90 days

75ms

Avg response, 30 days

Measured from multiple regions every 6 hours. Percentages count only checks that returned an availability answer — 21 days measured so far. A dash means that window does not yet hold enough measured days to publish a figure.

30-day history

21-day clean streak

Jul 25 Today

No downtime Partial Downtime Not measurable No data

Reachability by region

Each region runs its own request from a different part of the world. A service can be up for one continent and down for another, which is usually the first sign of a routing or CDN problem.

iad

94ms

DNS 0ms TCP 2ms TLS 7ms TTFB 83ms

jnb

70ms

DNS 2ms TCP 0ms TLS 5ms TTFB 64ms

lax

61ms

DNS 0ms TCP 1ms TLS 7ms TTFB 52ms

lhr

53ms

DNS 0ms TCP 2ms TLS 6ms TTFB 43ms

nrt

57ms

DNS 0ms TCP 1ms TLS 6ms TTFB 50ms

ord

61ms

DNS 0ms TCP 1ms TLS 7ms TTFB 53ms

sin

39ms

DNS 0ms TCP 1ms TLS 6ms TTFB 31ms

sjc

72ms

DNS 0ms TCP 4ms TLS 10ms TTFB 55ms

syd

46ms

DNS 0ms TCP 1ms TLS 6ms TTFB 39ms

What OpenRouter does

OpenRouter is an API gateway that routes requests to models from many providers through one endpoint and one billing relationship. Developers use it to move between models without rewriting integrations or holding an account with every vendor. That makes it a single dependency sitting in front of several others, so its own availability matters as much as any model provider's.

What an outage looks like

Chat completions fail at the gateway while the underlying model providers stay healthy, which is confusing because their status pages all read green. OpenRouter tracks the chat completions endpoint, the data API and the homepage separately, and lists account authentication through Clerk as its own component, so the dashboard can be unreachable while existing API keys keep working.

What to do about it

Check status.openrouter.ai, which reports the chat completions endpoint, the data API, the homepage and Clerk authentication, each with 90 days of uptime. Because OpenRouter sits in front of other providers, confirm the gateway before working through each model vendor. It has also posted an erroneous outage that it corrected within a minute, so re-read the page before acting on a brief red.

Is it down for everyone, or just you?

If this page says OpenRouter is up but it is not loading for you, the problem is between you and them. Run a check against any URL from all 18 regions to find out where it breaks.

Test it yourself

Related services

OpenRouter outage FAQ

Is the problem OpenRouter or the model provider behind it?

Start at status.openrouter.ai, because the gateway is the part you can check in one place. If its chat completions component reads healthy, the failure is more likely at whichever provider your request routed to, and OpenRouter can route to several. Sending the same prompt to a different model through the same key is a fast way to separate the two.

Can I still use my API key if the OpenRouter dashboard is down?

Often yes. OpenRouter lists account authentication through Clerk as a separate component from the API endpoints, so the sign-in flow for the web dashboard can fail while existing keys keep authenticating requests normally. If you only need to make calls rather than manage billing or keys, test a real request before assuming you are blocked.

Has OpenRouter's status page ever been wrong?

Yes, and it records the fact. In July 2026 it posted an incident titled erroneous outage status, resolved in under a minute, where it had briefly reported a problem that was not actually happening. That is a good argument for confirming a short-lived red against an independent measurement before rerouting production traffic.

Does OpenRouter going down mean all my models are unavailable?

If you reach models only through OpenRouter then yes, because it is the single endpoint in front of all of them. That is the trade-off for one integration and one bill. Teams that treat it as critical usually keep a direct account with at least one provider as a fallback, since the model vendors themselves are unaffected by a gateway outage.

How we measure this

We request OpenRouter's public endpoint every 6 hours from Fly.io regions across six continents — 8 of them answered the most recent check.
A region counts as down only when it gets no usable HTTP response. A 403 or 429 means the origin answered and refused us, which we report as blocked, never as an outage.
A single failing region is treated as probe noise. We only change the verdict when two consecutive cycles agree.
Response times average only the regions that actually served the page, so a timeout never inflates the number.
Where OpenRouter publishes an official status feed we read it too. An all-clear from the vendor can soften an unconfirmed degradation; a vendor-declared outage only worsens our verdict when our own checks corroborate it.

Get alerted when OpenRouter goes down

This page refreshes every 6 hours. Your own monitors run as often as every 30 seconds, from the same 18 regions, and tell you the moment something breaks.

Start Free Monitoring

Free plan available No credit card required