Cronjob.de features

Cronjob.de runs your cron jobs: we call your URLs on schedule, log your server's response and tell you when something breaks. And cron jobs that run on your own server, we watch with heartbeats. No cron daemon of your own, no server access needed, in operation since 2003.

Built-in AIpaste the instructions, the job is ready
Claude & ChatGPTconnect directly via MCP
Heartbeatsnotices when jobs on your server go missing
EU serversGDPR-compliant, German contract
New Heartbeats for jobs on your own server, alerts in Slack and Teams, access from Claude and ChatGPT and more See what's new
Set up with AI

Paste the instructions, we build the cronjob

Almost every set of instructions from a host, a shop or a CMS ends in a text that beginners cannot use. Just paste it here. We read out the address, the method and the schedule, ask if something is missing, and fill in the form with it. Nothing is saved until you submit it yourself.

  • Hoster instructions or crontab line – a curl or wget command works too
  • Asks instead of guessing – for the domain, say, when only a path is given
  • Says so when it will not work – for a command that only runs on your own server
  • Or just the schedule – „every second Tuesday of the month at 6:45" becomes 45 6 8-14 * 2
  • A suggestion, not a decision – nothing is applied without your check
  • You see when it runs – the next execution times are shown in plain text on the cronjob before you save

Try it now in the crontab generator →

Scheduling

Minute-precise, by click or cron expression

Pick minutes, hours, days, months and weekdays. If you know crontab syntax, just type it in. Both ways lead to the same result, and you immediately see when the job runs next.

  • Every minute down to once a year – any combination, including several times per hour
  • Folders – group cronjobs by project or customer
  • Test right away – start a job manually and see the response immediately
  • Password protected targets – credentials in the URL, also for .htaccess protection

Look up crontab syntax →

APIs

Not just call it, hand something over

You choose the method, send data along and set your own headers. That lets you drive someone else's API just as well as your own scripts – from handing an order to an ERP system to calling a service's webhook.

  • POST, PUT, PATCH, DELETE, HEAD – alongside GET, in every plan
  • Request body – JSON, form data or text, with the matching content type
  • Custom headers – an API key or a token, for example
  • Credentials in separate fields – any special characters in the password, and it never appears in the call log

Looking for an API to Cronjob.de itself? That exists as an MCP connection: manage cronjobs with Claude, ChatGPT or your own scripts. When creating a cronjob, the method sits right in front of the address, with headers and credentials below it.

Call log

You see what your server answered

A cronjob that seems to run but does nothing is the most expensive kind of failure. That is why Cronjob.de stores your script's response for every call, not just a checkmark. This cronjob log is available in your account at any time.

  • The last 20 runs – with date, time and the HTTP error in plain text
  • History over 30 or 90 days – runs and errors per day, plus every error period with a plain explanation, such as a block by Cloudflare
  • Up to 5,000 characters of output – from the latest run, older ones trimmed to 500
  • Duration and size – how long your script took and how much it returned

Every error is shown in plain words, with advice on what to do. The most common causes are explained in our troubleshooting guide.

Statistics

When a script slowly gets slower

The last 200 runs form a curve for duration and output size. That is how you spot an import that grew from 2 to 20 seconds, or an interface that has been returning empty responses since yesterday.

  • Failed runs as red dots
  • Time limit shown as a line in the chart
  • Date and value on hover

Included in the Unlimited plan.

Notification

You find out before your customer calls

If your server answers with an error or not at all, you get an email. We report the start and the end of an outage: one message when the cronjob fails and one when it works again. In between it stays quiet, even if it runs every minute.

  • An all-clear as soon as it works again
  • A second server verifies before we wake you
  • For all cronjobs or only for selected folders
  • Recipients of your choice, several possible
  • Optionally also in Slack, Teams, Discord or as a webhook to your own system

Email from Plus. Slack, Teams, Discord and webhook on the Unlimited plan.

Server monitoring

Is your site working – or just answering?

“200 OK” and yet the page says “Error establishing a database connection”. The add-to-cart button is gone because a plugin broke. For most monitoring services the page is fine – it answers, after all. We check every minute whether it works.

  • AI setup – Enter the address: the AI looks at the page and suggests how to tell it is working. It verifies every suggestion on the page first.
  • Error pages despite “200 OK” – PHP errors, WordPress without a database or stuck in maintenance, SQL errors, directory listings, empty pages
  • Texts that have to be there – “Add to cart” missing? Then something is wrong – even if the server looks happy. And texts that must not appear.
  • No flood of false alarms – we report the change, and a second, independent server confirms every outage
  • Alerts where you work – email, Slack, Discord or Microsoft Teams

We usually report three to five minutes after an outage starts. If you need alerts to the second, a specialised service suits you better – our strength is what gets checked.

Heartbeat

Monitor cron jobs that run elsewhere

Your server, n8n or Supabase runs the job – and checks in with us after every run. That is the heartbeat, a regular sign of life. If the check-in does not arrive, we alert you.

  • One line – curl in the crontab, PHP, Python, Node.js, n8n, Make, Supabase or Vercel
  • Failures right away – the exit code comes along; a call at the start measures the run time and catches stuck runs
  • One alert per incident – and the all-clear with the next check-in
  • Success check – for cronjobs we call: the script reports its end, or we check the reply for “OK” and for error messages (Unlimited)
  • 1 heartbeat with email free, 10 in Plus, 50 with every channel in Unlimited

Cron job monitoring: all lines and prices →

SSL monitoring

We see when your SSL certificate expires

Every time we call your cronjob we also see how long the certificate of that address is still valid. On request we let you know in good time – by email, to a different address than the cronjob alerts if you like.

Sounds like yesterday's problem now that Let's Encrypt renews itself? In practice things still expire: a certificate somebody bought by hand, a renewal that quietly fails, a subdomain nobody thinks about any more. An expired certificate stops payments and interfaces, and nobody tells you – until a customer calls.

  • Nothing to set up – we already know your addresses, a switch is enough
  • Twice, not once – once early enough to plan, and again shortly before it expires
  • Its own address – your tech people need to know, not necessarily accounting
  • Visible in your account – remaining validity and issuer of all your addresses on one page

Included in the Unlimited plan. Off by default – you decide whether you want to hear about it.

Maintenance mode

One switch while you work on your server

Installing an update, migrating, backing up the database – and your cronjobs keep running and keep failing. That costs error notifications, needless load and, in the worst case, half-finished imports. One click pauses everything, a second one starts it again.

The difference to switching things off by hand: we remember what was running. When you end it, exactly that state comes back – cronjobs you had deliberately switched off stay off. With 200 cronjobs you would otherwise have to tick them one by one and remember which ones they were.

  • All or part of it – the whole account, or just the folders whose system you are touching
  • Automatic end – 1 to 24 hours, so nobody forgets – paused cronjobs do not report anything
  • No error emails – while it is on, everything stays quiet
  • Schedules stay – nothing is rewritten, nothing is lost

Included in the Plus and Unlimited plans. You can end it by hand at any time.

Team logins

Several people, one account

Colleagues get their own login with their own email address and password. Everyone works under their own name, and when someone leaves, exactly their access goes – everyone else keeps theirs.

  • Two roles – members work with the cronjobs, admins also with invoices, plan and team
  • Invitation by email – a link valid once, revocable at any time
  • Activity log – who created, changed or deleted which cronjob, and when
  • One person, several accounts – agencies switch between their clients' accounts from the menu
  • No limit – included in the Unlimited plan
AI access (MCP)

Manage your cronjobs in conversation

Cronjob.de speaks the Model Context Protocol, MCP for short. Connect Claude, ChatGPT, Cursor or another assistant to your account and just ask – instead of clicking through lists.

  • Enter the address, confirm in the browser – or a fixed token for servers without a browser
  • You decide whether the assistant may only read or also make changes
  • Every connection can be removed in your account at any time
  • Included in the Unlimited plan

How to connect your assistant – step by step →

Cronjob.de infrastructure: six scheduler servers in Helsinki, Nuremberg and Falkenstein working from the same job list
Operations

Several data centres, so your job actually runs

Calls are spread across independent machines in Helsinki, Nuremberg and Falkenstein. Each takes its share of the jobs from the same list. If one fails, only part of the calls is affected.

  • Operated from Germany – German company, German support, invoice with VAT shown
  • GDPR compliant – servers in the EU, your data is not passed on
  • Two-factor login – optional, by email code
  • Running since 2003 – one of the first web cron services there was
  • Renewable electricity – according to their operator the data centres run entirely on hydropower, in Finland also on wind
Limits

The numbers at a glance

No asterisks, no hidden conditions. What applies in your plan:

Free Plus Unlimited
Cronjobs 5 unlimited unlimited
Smallest interval 5 minutes 1 minute 1 minute
Runtime per call 45 s 60 s 60 s
Output size 1 MB 1 MB 1 MB
Call log 5 runs, no response 20 runs 20 runs
History with error periods no 30 days 90 days
Runtime statistics no no 200 runs
Failure alerts no by email email, Slack, Teams, Discord, webhook
SSL certificate expiry warning no no yes
Server monitoring with AI setup and text checks no no yes
Team logins no no unlimited
AI access (MCP) no no yes
Confirmation that the jobs are still needed after 3 months without signing in, by email link not required not required

Every new account starts with 14 days of Unlimited. See pricing

In use

What our customers use Cronjob.de for

A cronjob here is always a call to a URL. What your script does behind it is up to you.

Shops and inventory

Export orders, sync stock levels, update prices, generate invoices.

Web hosting without cron

Many hosting plans offer no cron at all or only one call per day. WordPress, TYPO3 and Shopware need more.

Backups and maintenance

Back up databases, clear caches, tidy log files, delete temporary files.

Email and newsletters

Work through sending queues, send reminders, check subscriptions.

Interfaces and imports

Sync data between systems, read feeds, fetch exchange rates and weather data.

Reports and analytics

Collect key figures, generate daily and monthly reports and send them automatically.

Your first cronjob runs in two minutes

Email address, password, URL, schedule. That is all it takes. The first 14 days with every Unlimited feature, then continue for free.