Extend your gateway with AI Workspace

Requires AI Workspace

Everything on this page needs a running AI Workspace deployment. The gateway serves traffic without one.

The gateway serves traffic on its own. Connecting it to AI Workspace, the control plane for AI traffic across your organization, adds central management across every gateway you run.

What connecting the gateway adds

• One console across gateways. AI Workspace manages LLM providers, App LLM proxies, MCP proxies, and the credentials behind them for every gateway you connect.
• Central configuration. Configure an artifact in the console, then deploy it to one gateway or several.
• Artifacts in both directions. Artifacts you create through the gateway management API sync upward automatically and appear in the console as copies the gateway owns.
• Policies in one place. Attach guardrails, rate limits, and traffic controls to providers and proxies from the console.
• Traffic insights. Review token usage, latency, cost, and guardrail events across your applications and consumers.

Connecting a gateway doesn't make it depend on the control plane. If AI Workspace is unreachable, the gateway carries on and the sync catches up once the connection is restored.

Try it locally

You don't need a hosted deployment to decide whether this suits you. Get started with AI Workspace runs the control plane on your own machine, then walks you through creating a gateway and configuring a provider.

Connect your gateway

Four pages document the connection, each one suited to a different runtime and starting point. For the one that matches yours, see Connect the gateway to AI Workspace.