Skip to main content
The public Guardway management API is in active development. It will give you programmatic control over everything the dashboard exposes — gateways, providers, models, keys, teams, usage.
This section will be replaced with a full OpenAPI-backed reference when the API ships. In the meantime, the gateway itself already exposes a stable OpenAI-compatible inference API you can call today.

​
What’s available today

Inference API

Your gateway exposes /v1/chat/completions, /v1/embeddings, /v1/models, and more. It’s a drop-in OpenAI-compatible endpoint — anything that talks to OpenAI talks to your gateway.

MCP integration

Register MCP servers with the gateway to give models access to tools.

​
What’s coming

  • Gateway lifecycle (register, revoke, rotate credentials)
  • Provider CRUD with per-gateway attachment
  • Model inventory read / sync
  • Key management (issue, rotate, revoke, scope)
  • Usage and spend queries (same data as the dashboard, as JSON)
  • Webhooks for budget alerts and health events
Want early access? Email support@guardway.ai.