Capacity with a sense of direction.
Requests find eligible upstream keys by available capacity and budget. Shared limits and cooldowns stay part of the decision.
Capacity-aware routing
ONE KEY · TWO NATIVE APIS
Claude and OpenAI, through one considered connection. Native APIs. Clear usage. Room to build.
Your SDK. Your models. One RelayPine key.
A request enters the gateway.
Illustrative data. RPM shown; other limits assumed available. Budgets are local caps, not provider balances. No API calls.
BUILT AROUND YOUR STACK
THE FAMILIAR, CONNECTED
Use the OpenAI and Anthropic libraries you already know. Update your key and base URL, then get back to building.
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.RELAYPINE_API_KEY,
baseURL: "https://relaypine.com/openai/v1",
});
const response = await client.responses.create({
model: "MODEL_ID",
input: "Hello from RelayPine",
});Set RELAYPINE_API_KEY to a key created in the console. Replace MODEL_ID with an enabled model ID from Models & rates. API access is limited during private preview.
THE DETAILS, TAKEN CARE OF
The infrastructure behind every request, brought together in one workspace.
Requests find eligible upstream keys by available capacity and budget. Shared limits and cooldowns stay part of the decision.
Capacity-aware routing
Follow token usage, request status, and latency across both providers in one place.
Input · Output · Cached tokens
Give each application its own key, model access, and rate limits. Keep your projects in control.
Separate keys · Defined limits
Sign in with your email to create a workspace and connect your applications.