Bring your keys. Get unified control.

Plug In. Talk to Any Model or Agent. Track It All.

Ship AI features faster with a single endpoint—no backend mess, retries, or guesswork. Logs, costs, and latency tracking included.

Use Cases

  • Cross-model orchestration (GPT, Claude, Gemini)
  • Smart routing with fallback & cost-aware retries
  • Private RAG & custom agent pipelines
  • Plugin-driven workflows & marketplace integrations
  • End-to-end observability & real-time metrics
  • Hosted billing & usage metering

Why Developers Choose Relay

Universal RelayPlane API

One POST /relay call to 50+ models, private RAG endpoints, or custom agent chains.

Open-Source MCP Router

Drop-in Rust router for multi-agent choreography—fully MIT-licensed on GitHub.

Python-A2A Adapter

Publish and pip install our adapter for seamless LangChain & AutoGen integration.

Real-Time Metrics

Track cost, latency, and usage in your dashboard or via our REST logs API.

🚧 What's Coming to RelayPlane

We're just getting started. Here's a glimpse of what's ahead as RelayPlane evolves from a developer-friendly API to a full AI control plane.

✅ In Development

  • Lightweight JS & Python SDKs (3 lines of code).
  • Agent-chaining playground for rapid prototyping.
  • Unified logging + observability across all calls.

🔄 Coming Soon

  • Optimize: built-in caching & smart cost routing.
  • Plugin SDK v0: write custom agents & tools.
  • Agent Marketplace: buy, sell, and share workflows.

🏢 On the Horizon

  • Private cloud & enterprise deployments.
  • SOC-2 Type I compliance.
  • Advanced analytics for team ops.

Simple pricing that scales with you.

Free

Ideal for indie devs—pay only for usage.

$0/mo
  • Hosted metering, dashboard & logs
  • Up to 5k calls/mo
Join Waitlist
Most popular

Maker

100k calls included, then pay-as-you-go.

$15/mo
  • Unlimited API keys
  • Email support
Join Waitlist

Growth

Scale with predictable unit economics.

$0.02/1k calls
  • Volume discounts available
  • Dedicated uptime SLA
Join Waitlist

Agent Marketplace

Coming soon: Build, discover, and monetize AI agent pipelines.

RelayPlane will feature paid plugin ranking, revenue sharing, and collaborative workflows.

Run Your First Call (Quickstart)

example.js
import { relay } from "@relayplane/js";

async function main() {
  const res = await relay({
    to: "gpt-4o-mini",
    payload: {
      messages: [{ role: "user", content: "What's the capital of France?" }]
    }
  });
  console.log(res.body); // { "message": "The capital of France is Paris." }
}
npm install @relayplane/js

Frequently Asked Questions

What is Relay?

RelayPlane is the AI control-plane: one API for model calls, agent chains, cost/latency insights, and hosted billing. Join the waitlist to get started.

When can I use it?

Private beta is in progress. Join the waitlist for early SDK access.

Do I still need OpenAI / Anthropic keys?

Yes: RelayPlane brokers your own keys so you keep full cost & usage ownership.

How does billing work?

Freemium + pay-as-you-go metered billing—no surprises.

Latest from the Blog

Insights, tutorials, and updates on AI infrastructure

View All Posts
RelayPlane vs. LangChain: Infra Layer vs. Dev Framework
Product
LangChain helps you build agent apps. RelayPlane helps you run them—reliably, observably, and across any model or agent. Here's how they differ.
May 30
3 min read
RelayPlane vs. LiteLLM: More Than Just a Model Router
Product
LiteLLM gives you access to multiple LLMs with a single API. RelayPlane goes further—handling retries, failover, agent-to-agent workflows, and full observability. Here's how they compare.
May 29
4 min read
Introducing RelayPlane: The Universal Control Plane for AI Models and Agents
Product
Today we're launching RelayPlane, the missing layer between your app and every AI model or agent you need. One API. Total control. Zero backend duct tape.
May 26
3 min read

Building with Industry Leaders

Anthropic

Anthropic

OpenAI

OpenAI

LangChain

LangChain

Vercel

Vercel

Ready to connect your AI agents?

Join the early access program and be among the first to experience seamless agent orchestration.