Delay is a hosted cron job service for developers. It calls your endpoints on a schedule or when your code asks, and tells you when they fail.
Almost every app ends up with scheduled work: a nightly sync, a weekly report, a cleanup of old records. The usual home for it is a crontab on some server. It works until the server is rebuilt, a deploy drops the crontab, or a script starts exiting early. Cron keeps quiet either way, and you find out when a customer does.
Delay moves the schedule off your servers. Your app keeps the code that does the work, behind an ordinary HTTP route, and Delay calls that route on time and keeps a record of what came back. When something breaks, you hear about it the same day.
A cron is a URL, a method, optional headers and a body, and a cron expression in the time zone you choose. Delay makes the request on schedule and saves the status code, the response body and how long it took. If a cron fails as many times in a row as you allow, your team gets an alert in Slack, Discord or email, and another one when it recovers.
For work that doesn't run on a fixed schedule, send a job to a queue with one API call. It runs right away or up to 24 hours later. If it fails, it retries on the delays you set for that queue, and if it never succeeds it waits in the dead letter queue until you replay it.
A team has admins and members. Everyone sees the same crons, queues and run history, and alerts go to the whole team, so nobody needs to share a login to find out why last night's job failed.
Over the API you can create, change, pause, run and delete crons, schedule and cancel jobs, and read every run with an explanation of what went wrong. AI agents such as Claude Code can do the same through Delay's MCP server, with a token you can revoke at any time.
Read the API reference