Integration Guide

Lambda Labs + Meridian

Deploy GPU-accelerated inference workloads on Lambda Labs infrastructure while Meridian handles licensing, metering, and customer identity — so you ship features instead of billing code.

Get Started Lambda Labs →

Architecture Overview

Meridian sits between your customers and your Lambda Labs GPU fleet. Customers authenticate through Meridian, receive short-lived tokens, and your inference API validates those tokens before dispatching work to Lambda instances.

Customer Auth

Meridian issues Ed25519-signed license tokens tied to hardware fingerprints and subscription tiers.

Token Validation

Your API edge validates tokens via Meridian SDK before accepting inference requests.

GPU Dispatch

Validated requests hit your Lambda Labs instances running vLLM, TGI, or custom runtimes.

Quickstart

Install the Meridian SDK in your API layer:

npm install @getnimbus/meridian-sdk

Validate a license token before forwarding to Lambda:

import { validateToken } from '@getnimbus/meridian-sdk';

const result = await validateToken({
  token: req.headers['x-meridian-token'],
  tier: 'pro',
});

if (!result.valid) {
  return new Response('Unauthorized', { status: 401 });
}

// Forward to Lambda Labs instance
await fetch('https://your-lambda-instance:8000/v1/chat/completions', {
  method: 'POST',
  headers: { 'Content-Type': 'application/json' },
  body: JSON.stringify(payload),
});

Recommended Tiers

Map Meridian license tiers to Lambda Labs GPU types. Customers upgrade through Meridian — your infrastructure scales automatically.

Tier	Lambda GPU	Rate Limit	Price
Starter	1× A10	60 req/min	$49/mo
Pro	4× A100	600 req/min	$299/mo
Enterprise	8× H100	Unmetered	Custom

Ready to ship?

Connect your Lambda Labs account to Meridian and start metering GPU inference in under 10 minutes.

Open Dashboard Full Docs