AI Gateway entities

Register upstream LLM providers with authentication and configuration. Reusable across AI Models.

Define how to reach and interact with a specific LLM, including routing, load balancing, LLM capabilities, and AI Policies.

Expose, observe, and secure autonomous agents through AI Gateway for tool-calling and decision-making workflows.

Proxy MCP servers to secure, govern, and observe tool-calling workflows.

Control security, rate-limiting, and guardrails across AI Models, AI Agents, and AI Consumers.

Manage downstream client authentication and access control to your AI APIs.

Group AI Consumers together to apply shared rate limits, policies, and access rules.

Securely store and reference API keys, tokens, and credentials across all AI Gateway entities.

Configure inbound authentication for AI Consumers calling AI Models. Supports API key and OIDC token authentication.

Authorize data planes to connect securely with mTLS certificates.

Store certificates and private keys used to terminate and originate TLS connections.

Store public certificates from trusted Certificate Authorities to verify the validity of AI Certificates.

Map hostnames to an AI Certificate so data planes present the right TLS certificate to each client.