Host your AI Agents on Steamship

Auto-scale your agent stack

Scale agents from 0 to 1000s of requests per second automatically. No need to worry about infrastructure.

Zero configuration

Auto-scaling is built into the SDK and requires no additional code

Async by default

Every operation is an async task under the hood

Managed lifecycle

Resources like vector DBs and tuneable models are auto generated for your agents.

Deploy once, reuse everywhere.

Create multiple instances of your deployed agents to scale across users without designing a multi-tenant architecture.

Isolated Data

Each instance manages its own isolated data context: vectors, generations, uploads, and chat history.

Private APIs

Each instance publishes private APIs for callbacks, management, and integrations.

Segmented Billing

Billing is tied to agent instances, not the code that defines them. Perfect for developers with many clients.

Easy deployment

Use our CLI to deploy your agents to the cloud in seconds.

CLI Deployments

Deploying your AI Agents to Steamship is as easy as running a single command: ship deploy. We take care of the rest.

Versioning

Version your agent deployments. Simply run ship deploy --version 1.0.0 and we will deploy your agent to the specified version.