Watch every API you depend on, even behind a login.
ThirdWatch signs in, keeps the token fresh and checks your endpoints as often as every minute. When one fails, it checks again before alerting your team, and tells them when it's back.
Starter plan · No credit card · Cancel anytime
- UpCheckedHTTP 200 · 142 ms
- TokenToken refreshedOAuth2 · valid for 60 more minutes
- DownHTTP 503Checked again 10 s later: still down
- AlertTeam alertedIncident opened · Slack #ops
- UpBack up · 156 msDown for 4 min · Slack #ops told
It signs in. Your checks no longer stop at the door.
Most monitors give up at the first 401. ThirdWatch logs in the way your API expects and handles the token from start to finish.
- 1
Signs in
OAuth2 with client credentials or a password, your own login request, an API key, a bearer token or Basic auth. Secrets are stored encrypted.
- OAuth2
- Custom login
- API key
- Bearer token
- Basic auth
- 2
Renews the token
Reads the token where your API returns it, with its expiry, and renews it a minute before it expires, by the refresh token or a new login. If your API still answers 401, ThirdWatch renews it and checks again.
- 3
Checks with it
Every check and every scenario step carries the token where you said: a header or a query parameter.
Figures you can show your customers.
Uptime and response times are counted from every check, from the last 24 hours to a full year depending on your plan. See the live demo.

Test the journey, not just the endpoint.
Chain calls the way your users make them: sign in, create, read, delete. A value from one step feeds the next, and assertions say what success means.
Scenarios come with Pro and Enterprise, scheduled as often as every 15 minutes on Pro and every 5 on Enterprise, or run by hand.
- 1POST/auth/login200 · 212 msExtracts
token - 2POST/orders201 · 340 msSends
{{token}} - 3GET/orders/{{order.id}}200 · 98 msAsserts
status = "pending" - 4DELETE/orders/{{order.id}}204 · 120 msCleans up after itself
Secrets stay masked in the run history: Authorization: Bearer ••••7890
One alert, when it's real.
A failure is checked again before anyone is paged. Then each channel of your rules gets one message with what failed, and another when it's back, with how long it lasted.
- Where alerts go
- Slack · Discord · Microsoft Teams · Webhook
- No flood
- A quiet period per rule keeps a flapping API from flooding the channel; an outage that begins during it is announced as soon as it ends.
- Certificates
- A TLS certificate close to its expiry is announced 14, 7, 3 and 1 day before.
And the rest of the watch.
- OpenAPI import
- Drop a Swagger 2 or OpenAPI 3 file: one monitor per endpoint, with path parameters filled from the spec's examples.
- Response rules
- Expected status, keywords, JSON fields, strict JSON and response time, on any method, with your headers and body.
- Monitor groups
- By service or by client, with their own alert rules and default settings for new monitors.
- Incidents
- Opened, acknowledged and resolved, each with its timeline. Resolve one while the API still fails, and it opens again.
- Team and roles
- Owner, Admin, Editor and Viewer, invitations by email, and several organizations under one account.
- History
- 30 days of checks on Starter, 90 on Pro and a year on Enterprise.
- VPN routingQ1 2027
- Reach internal APIs through your VPN, on Enterprise.
Three plans, a trial to start.
Starter watches your first APIs for €11.99 a month, after 14 days free without a card. Pro adds multi-step scenarios, Enterprise checks every minute.
Starter
14-day free trial€11.99/month
20 monitors checked every 5 minutes, 3 alert channels, 30 days of history.
Pro
€25.00/month
50 monitors every 2 minutes, 20 API scenarios, unlimited channels, 90 days of history.
Enterprise
€99.00/month
200 monitors every minute, 100 scenarios as often as every 5 minutes, a year of history.
Someone has to stay up at night. Let it be ThirdWatch.
Starter plan · No credit card · Cancel anytime