Backend code that runs itself, on demand or on schedule.
Serverless functions triggered by a request, a schedule, or a webhook, running in 300+ cities.
Part of a Solution. Built, released and secured together with the rest of your app: one place, one release.
What it does
What you get with Ambassadors
Pay for work, not waiting
You’re charged for the time your code is computing, not while it waits on an API or database.
No cold starts
Functions answer instantly, even the first request after a quiet night.
Runs near your users
Deploy once; it runs in 300+ cities, close to every request.
Launch-day ready
Scales from zero to millions of requests with nothing to pre-provision.
Runs on your schedule
Cron-style schedules for reports, syncs and clean-ups.
Connects your world
Each function has its own HTTPS endpoint for webhooks, with secrets stored encrypted.
Use cases
What people build with it
A payment webhook
Receive the event, update the order, send the receipt, stop. You pay for the milliseconds of work.
The Monday-morning report
A schedule runs it at 7am, it reads the database and emails the result.
Two systems kept in step
An hourly sync between your app and the tool your team already uses.
The API behind your app
A function per endpoint, answering in the city nearest the request.
Not a fit for
Ambassadors wake, do the work, and stop. If you need a process that stays up, this is the wrong tool, and no configuration will change that.
- Game servers, or bots holding many persistent socket connections.
- Video transcoding pipelines.
- Data jobs measured in hours rather than seconds.
- Anything that must stay resident in memory between requests.
For AI work that has to remember and keep going, use Agents (in preview). For a container that must stay up, a container platform such as Fly.io, Railway or Render will serve you better, and we would rather you succeed there. About Agents.
How it’s charged
Metered per execution, by CPU time, drawn from your plan’s credits today. Time spent waiting isn’t charged.
See Host pricingThe technical detail
Commands, limits and step-by-step setup for Ambassadors live on the Docs site.
Read the docsQuestions
Do Ambassadors have cold starts?
No. Functions answer instantly, including the first request after a quiet night, and run in 300+ cities close to each request.
Am I charged while my function waits on a database or API?
No. Ambassadors are metered by CPU time, the time your code is actually computing. Time spent waiting on a network call isn’t charged.
Can an Ambassador run a long-running process?
No. Ambassadors are request- and schedule-scoped: they wake on a trigger, do the work and stop. For stateful AI work use Agents (in preview); for a container that must stay up, use a container platform such as Fly.io, Railway or Render.