RelayRouterOne gateway.
Claude, GPT and Gemini.
Production access to Claude Opus 4.8, Claude Fable 5, GPT-5.5, Gemini 3.5 and more through one gateway that speaks the OpenAI, Anthropic and Gemini APIs natively. Keep your SDK, swap the base URL.
Models from the world's leading AI labs
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.RELAYROUTER_API_KEY,
baseURL: "https://relayrouter.io/v1", // the only change
});
const res = await client.chat.completions.create({
model: "claude-opus-4-8",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(res.choices[0].message.content);One API
One API. Every top model.
No more juggling SDKs or API keys. Production-ready models behind a single OpenAI-compatible endpoint. Change one line of code.
- ✓Claude Opus 4.8, Claude Fable 5, GPT-5.5 and Gemini 3.5 in one catalog
- ✓Three protocols: OpenAI, Anthropic and Gemini compatible endpoints
- ✓Mainstream model groups are priced 30% or more below official list prices, zero platform fee
- ✓Pay by credit or debit card via Stripe
Integrate in minutes
Keep your SDK. Change one line.
OpenAI SDK, Anthropic SDK or raw HTTP: set the base URL, swap your key, done.
import OpenAI from "openai";
const client = new OpenAI({
apiKey: process.env.RELAYROUTER_API_KEY,
baseURL: "https://relayrouter.io/v1", // the only change
});
const res = await client.chat.completions.create({
model: "claude-opus-4-8",
messages: [{ role: "user", content: "Hello!" }],
});
console.log(res.choices[0].message.content);AI Agents
Coming soonA dedicated MCP server is on the way.
Until it ships, any MCP-capable or OpenAI-compatible agent tool can already use RelayRouter today: point its base URL at https://relayrouter.io/v1 and supply your API key.
{
"mcpServers": {
"relayrouter": {
"command": "npx",
"args": ["-y", "relayrouter-mcp"],
"env": { "RELAYROUTER_API_KEY": "your-api-key" }
}
}
}Compatibility
Works with your existing stack.
Any OpenAI-compatible client, SDK, or framework.
Use cases
Built for every AI workflow.
App development
Build AI-powered products without locking in to a single provider. Switch models as they improve, no code changes required.
AI agents & automation
Route tasks to the right model: fast models for drafts, frontier models for decisions. One API, any combination.
Research & evals
Compare outputs across models with one API call. Benchmark quality, latency and cost across every available model.
Why RelayRouter
Everything you need, nothing you don't.
Claude-first catalog
The full Claude line (Opus, Sonnet, Haiku) next to GPT and Gemini. Change model with one parameter.
Three native protocols
OpenAI, Anthropic and Gemini compatible APIs on one base URL. OpenAI SDK, Anthropic SDK and Gemini clients all work as-is.
No platform fee
Pure per-token pricing. Mainstream groups 30%+ below official rates. No surcharge, no subscriptions, no minimums.
Card payments via Stripe
Pay with credit or debit card through Stripe checkout. Top up any amount in seconds.
Zero failed-request charges
Errored or failed API calls are never billed. You only pay for tokens successfully processed.
Real-time analytics
Usage, latency, cost and throughput per model, all in one dashboard.
FAQ