What we shipped.

On a phone, a monitor’s page is now one compact column, about a third shorter. What matters during an outage comes first: the status, the error and the buttons to acknowledge and open the incident. Response time phases, results per location and the check rules sit under More about this monitor, and the incident list shows the five newest with Show all. On a computer the page is unchanged.

On small screens, alert channels now show a channel’s delay, such as “after 5 min”, in full, and Send test says it is sending while the test is on its way.

Export CSV on the incidents page downloads every incident your workspace has kept, newest first: the monitor and its target, when it started, was confirmed and was resolved (in UTC), the duration, the cause, the check locations, who acknowledged it and a link to it. Use it for an SLA report, a quarterly review or your own spreadsheet. However long the history, the file starts downloading at once. Available on every plan.

The REST API’s paged lists, incidents and a monitor’s checks, now return a nextCursor. Pass it as cursor to get the next page. Rows that share a time, such as checks from several locations in the same second, are never skipped or repeated, and hasMore is exact. before works as it did.

A status page can now be opened by the people you list instead of a shared password. Under Settings › Email access, add single addresses or a whole domain such as client.com. A visitor on the list enters their email and gets a sign-in link that works for 30 minutes; anyone else gets the same answer and no email, so the list can’t be guessed. Take someone off the list and they are shut out on their next visit.

Email access works with or without a password, and the list never reaches the visitor’s browser. Private pages are never indexed and take no subscribers. Available on every paid plan.

An SMS channel can now go to Whoever is on call in a schedule, like a push channel. Each person adds and confirms their own mobile number under Account, and the text goes to the number of the person on call when the alert is sent, overrides included. The on-call page warns about anyone in the rotation who hasn’t confirmed a number yet. SMS uses credits as before: included on Professional and Business, packs on every paid plan.

Moving to a smaller paid plan now works like moving to Free. The change starts when your current period ends. Then whatever doesn’t fit the new plan’s limits is paused, the oldest stay on, and nothing is deleted. Upgrade again and it resumes by itself, or delete what you no longer need and resume the rest. The workspace owner gets an email listing what was paused.

Every domain monitor now has its own DMARC report address. Add it to your DMARC record’s rua= list, next to the addresses you have, and the daily reports from Google, Microsoft, Yahoo and other mailbox providers arrive in Uptimecheck instead of as XML attachments in someone’s inbox.

The monitor names every sender of mail in your domain’s name, such as Google Workspace, Microsoft 365, SendGrid or an unknown server, shows whether it passes SPF, DKIM and DMARC and how to fix it, and tells you when it is safe to move to p=quarantine or p=reject. You are alerted when a new unknown sender fails DMARC, or a sender that passed starts failing. Available on every plan, Free included.

Slow alerts no longer need a number you have to guess. With Slower than usual, Uptimecheck learns each monitor’s usual response time over its first two days of checks and keeps learning from the last 14 days. When three checks in a row are much slower, about three times usual and at least 400 ms more, you get a slow alert that names the usual time, and another when it is back near normal.

New monitors start with it. Monitors you already have keep their fixed limit, or stay off, until you change them under Slow alerts. The monitor’s page shows the usual time and the limit in force, and the response time chart draws it. Through the API, set slowAuto.

Setting up a workspace on a phone? The alerts step now has Get alerts on this phone: one tap turns on push notifications for that phone and sends a test, so the first real outage reaches you there. Trials that have a monitor but have never received an alert get one email after about two days with a reminder to send a test alert.

Every password field has a Show password button. An email sign-in link that has expired or was already used now says so and lets you send a new one, instead of opening an empty sign-in page. The address you typed carries over to Forgot password? and back after the reset. The help center and Contact support are in the account menu and in the command menu (⌘K).

The app, its emails and this website now speak English, German, French, Spanish, Italian, Dutch, Polish, Swedish, Danish, Norwegian and Finnish. Each person picks a language in their account, otherwise the browser’s is used, and alerts to shared channels such as Slack follow the workspace owner’s language. The API, the MCP server and the admin stay in English.

Uptimecheck’s own status is public at status.uptimecheck.com. It is checked every minute from Cloudflare, outside our own servers, so it stays up when we don’t. It shows monitoring and alerts, cron job pings and server agents, and this website, with 90 days of history and recent incidents in UTC. Machines can read /status.json.

Website, API, port and ping checks now rotate between Stockholm, London and New York, and a failed check is confirmed from another location before an incident opens. A network problem near one location no longer looks like an outage. Your data, the app and backups stay in Stockholm and in the EU.

After the 14-day trial, a workspace that doesn’t pick a paid plan moves to Free instead of stopping: 5 monitors checked every 5 minutes, 2 cron jobs, 1 domain, 1 DNS monitor, 1 status page and 3 alert channels, with 90 days of history. Commercial use is fine. Anything over the limits is paused, never deleted, and starts again when you upgrade.

Every service page on Is it down? now has Email me when it goes down. Confirm your address once and we email you when the service has been down in our checks for 5 minutes, or at once when its own status page confirms it, and again when it is back. It is free, needs no account, and every alert has a one-click unsubscribe link.