Deploy

By the end of this page you will have deployed your AI Agent — either into an existing deployment or a new one — and know where Scheduled and API modes are enabled.

Prerequisites

• A version of the agent in PRODUCTION (Publish).
• Admin role.

Why deploy

Publishing makes a version live; deploying makes the agent reachable and unlocks the deployment-only modes. A deployment gives your agent a URL, authentication, and channels — and it’s where Scheduled and API modes are configured.

Deploy from the editor

Open the Deploy menu

In the editor header, click Deploy. The dropdown is enabled once the agent has a production version and the project has a short ID.

Choose new or existing

• New Deployment — start the deploy wizard to create a deployment for this agent.
• Add to: an existing ready deployment — listed by app slug and deployment type.

Complete the deploy wizard

The wizard walks through target → agents → configure → review: choose or name the deployment, confirm the agents included, set per-agent configuration and authentication (including API keys), and review before deploying. It then provisions and reports when the deployment is ready.

For the deployment lifecycle in depth — types, URLs, agents, and management — see the deployment guides:

• Deployments overview
• Production deployments
• Managing agents
• Custom domains

Enabling Scheduled and API modes

Once the agent is deployed, its Scheduled and API modes can be used:

• Scheduled — run the agent on a cron schedule with a fixed goal and inputs, for recurring autonomous jobs. See Scheduled runs.
• API — call the agent as a stateless REST endpoint, with a per-minute rate limit and API-key authentication. See API access.

Common issues

• Deploy is disabled — The agent needs a production version first. See Publish.
• The deployment serves old behavior — Confirm the agent is included in the deployment and that your latest version is in PRODUCTION.
• API calls return 401 — Use a valid API key or token for that deployment.