Skip to main content
Write with AI. Run locally. Deploy to Xano. This section is the last step: getting the backend your agent wrote, and you ran on the Xano Engine, into production on Xano. Production always runs a release: a stored, named copy of a backend that came up and answered on Xano’s cloud. You cut it once, and the same release lands on your workspace, a tenant, or a second workspace, as often as you need.
The Engine and the ephemeral are the same backend, one on your machine and one in Xano’s cloud. A release is cut from the cloud one, which is why the ephemeral step is there.
No Xano account yet? Connect to Xano takes the quickstart project to its first deploy to an ephemeral. Start there, then come back.

​
Start with where code goes

Every command in this section writes to a different place, and each one writes differently: an ephemeral is replaced, a workspace gets a new Xano branch, a tenant is replaced, a merge adds and updates. Where code goes puts all of them in one table. Read it before you run anything against a real target.

​
Pick your path

  • One workspace, which is production. Cut a release and promote it: Releases, then Promote and go live.
  • Tenants for dev, stage and prod. Cut a release once and deploy it to each tenant: Tenants.
  • Several workspaces or instances. Copy a release across, then promote it there: Across workspaces.
  • No releases. Merge a local build straight into the workspace or a tenant, with a preview first: Merge path.
Two pages apply whichever path you take. Source of truth settles whether your repo or your workspace holds the real backend. Shipping workflow lays the whole flow out for a team, from a git branch to production. Going back is covered under Govern: Rollback and recovery.

​
Next

Where code goes: what each command writes, and where.