Secure AI Gateway
A governed front door for every model call your org makes. One key, one endpoint, full audit trail — instead of a loose API key copy-pasted into every service that needs a model.
Key. Call. Govern.
01
Key
Use the same NEXORA API key your org already has — no separate Gateway key.
02
Call
POST to /v1/chat with a messages array, same shape as a normal chat completion.
03
Govern
Every call is authenticated, rate-limited, entitlement-checked and logged.
What’s real today
- →One upstream provider today, proxied with your org's own key never exposed to the caller
- →Every call authenticated, rate-limited and entitlement-checked the same way as every other NEXORA endpoint
- →Usage logged to your organisation's Console → Usage view
- →A live playground in Console to try it without writing code first
One configurable upstream provider today; multi-provider routing and a model policy engine are on the roadmap — see Concepts.