Overview

Relixr is a single, OpenAI-compatible endpoint that routes your requests across AI providers with routing, guardrails, and unified billing built in.

Relixr sits between your app (or your agents) and the AI providers you use. You send one request to Relixr's gateway, and it forwards it to the right model — Anthropic, OpenAI, Google, Mistral, Groq, and more — using the same request and response shape as the OpenAI API. Nothing new to learn on the client side.

Tip:

Already using the openai SDK? Change one line — the base_url — and you're on Relixr. Everything else stays the same.

Why Route Through a Gateway

  • One key, every provider. Stop juggling separate SDKs and credentials for each model provider.
  • Routing you control. Send cheap traffic to a fast, low-cost model and reserve expensive models for what needs them — declaratively, without changing your code.
  • Guardrails at the edge. PII redaction, prompt-injection checks, geo rules, and spend limits are enforced before your request ever reaches a provider.
  • Transparent cost. Every response tells you exactly what it cost, which model handled it, and why.
  • Built for agents. Autonomous agents get their own scoped keys, budgets, and usage attribution — not an afterthought bolted onto a human-facing dashboard.
Your apprequestRelixr gatewayguardrailsroutingAnthropicOpenAIGoogle, more…response + cost + route
A request travels from your app through Relixr's edge gateway to the selected provider, and the response — plus routing and cost metadata — travels back the same way.
Next steps
Create your first API key
Two minutes in the dashboard.
Make a request
Python, Node.js, or curl.
Agents
Give autonomous agents their own scoped identity.
Create your first API key