What changed in PingWise and from when. Every entry here is live for every account unless it says otherwise. Atom feed.
The fastest interval is now 10 seconds, from 15. It is available on hosts under a verified domain, on accounts that have bought credits at least once. At that interval a monitor costs 259,200 credits a month. Across all accounts a single host is still checked at most 12 times a minute, so the fastest intervals can be unavailable for a host that several accounts monitor. The home page calculator and Intervals and credits have the numbers.
When we answer a ticket you get the reply by email, and since 18:26 UTC today you can simply answer that email: your reply joins the ticket as if written in the panel, the ticket reopens, and we are told. The reply address is unique to the ticket (support+…@reply.pingwise.com) and never works as a login. Attachments sent by email are not added to the ticket; put them in the panel.
Also new: Request a feature in the user menu now has a building status, and requests can be marked public so they appear on the roadmap. The Privacy notice was updated the same day to say that Postmark, our email provider, also receives the mail you send to our support addresses, and that mail matching no ticket is deleted within 30 days.
Since 17:09 UTC a status page has a language: English, Greek, or auto, which follows the visitor's browser language and never their location. Everything on the page, its history, maintenance and feed, and the subscriber emails follow it. The status words are Λειτουργεί / Μειωμένη απόδοση / Μερική διακοπή / Διακοπή / Συντήρηση.
Earlier the same day, a round of settings people asked for after comparing with other services:
The status pages docs describe all of it.
Since 15:36 UTC every public status page has:
badge.svg: an SVG badge of the page's state, or of one entry with ?entry=<id>, with theme, style and size parameters. Cached 60 seconds, no cookie, no script.status.json: the page as data, with each component's id, state, 30- and 90-day uptime, when its current state began and its last incident. CORS is open, so a browser on any site can read it.Status pages also moved to status.pingwise.com (old status.uptimebox.io links keep working), and custom domains point their CNAME at pages.pingwise.com. Details in Badges, status.json and embedding.
Two changes on the Incidents page tonight:
The Incidents page also opens on the open incidents now, with the history under a Log tab, and the sidebar scrolls on its own on small screens.
Heartbeats. A heartbeat monitor uses one credit per expected check-in, the same as a scheduled check at that interval: a job that must call in every 5 minutes costs 8,640 credits a month. The pricing section says so from today, and the panel, the API's credits_per_month, import, bulk change and quick add show that cost instead of 0. The 650,000 free credits a month cover it for most accounts. Until today the panel showed heartbeats as free while the ledger charged them; no account on PingWise had a heartbeat monitor or had spent bought credits in that period, so nothing was owed to anyone.
Webhooks. New and changed webhook channels must use https, as the help text always said; an alert carries details about your site and should not travel in the clear. Webhook channels added over plain http before today keep working and are marked «http · needs https when changed» on the Alerts list. ntfy topics may still use http. Webhooks has the rules.
When a Cloudflare challenge or block answers the probes, a monitor shows blocked, a warning rather than an outage. Until 15:36 UTC today every scheduled round that met the challenge again started a fresh set of confirmation rounds, which was noise for the site and for our fleet. Now a monitor that is already blocked stays blocked quietly until the answer changes; the state, the alert and the panel are the same as before. The Cloudflare check tells you in advance whether a WAF treats the probes this way, and the probe IPs are public so you can allow them.
The same deploy added per-key API limits: a key can be given higher read, write or check-now limits on request, and the rate-limit headers then show that key's own numbers.
Since 08:52 UTC PingWise has a public API, version 1, at https://app.pingwise.com/api/v1. Keys are made under Team → API keys by an owner, read-only or read-and-write, and sent as a bearer token. It covers monitors (create, change, pause, resume, delete, check now), incidents, groups, status pages, the account's credits, and for resellers customer accounts and invitations. Lists use cursors, errors carry the same messages as the panel, and every response has rate-limit headers (120 reads, 20 writes and 10 check-now calls per key per minute). The API docs are generated from the same OpenAPI file the server is tested against, and API keys explains the keys.
Every monitor also gained a notes field of up to 2,000 characters, for what the person on call needs: the server, the contact, the usual cause. Notes are included in down, partial and reminder alerts and shown on the monitor page, never on status pages.
Since 19:02 UTC the Monitors page has a one-line Quick add box. Paste one address or twenty, with or without a scheme: ports 80, 443, 8080 and 8443 become HTTP monitors, 25, 465 and 587 SMTP checks, anything else a TCP check. Each becomes a draft with its cost in credits, confirmed before it is created. On accounts with the AI features turned on, the box also reads plain text such as "the shop and the API every minute, the blog every 5".
The free monthly allowance is now 650,000 credits, up from 100,000: enough for fifteen monitors checked every minute, or seventy-five every five minutes, at no cost. It renews on the first of each month and is spent before any credits you bought. The same day the voting fleet grew to 15 probes on 5 providers across 6 continents; a down verdict needs 10 of them, on 3 continents and 3 providers. The probe list and the transparency page show the fleet.
PingWise, the uptime monitor we built because our own servers needed one we could trust, opened to everyone at 17:52 UTC. We are a team of five, working remotely from Greece, Sweden, Serbia and North Macedonia, with fifteen years of hosting and projects of our own behind us. Prepaid credits instead of plans and seats, probes on six continents, and an alert only when continents and providers agree. The same day brought the TTFB check on HTTP monitors (a slow threshold measured as time to first byte, server time or full response) and the full set of alert channels: email, Telegram, Slack, Discord, Microsoft Teams, Google Chat, PagerDuty, ntfy, Pushover and webhooks.
Three things that make a hundred monitors manageable:
Import from another service has the CSV columns and the rules.