AI Gateway & Wallet Infrastructure
Route AI traffic, enforce prepaid usage, and monetize AI apps with one gateway.
AI usage is hard to monetize safely
Most teams ship AI features fast — then spend a year building the billing, quotas and routing they didn't plan for.
Unpredictable AI costs
Token spikes blow through margins overnight.
Flat-rate pricing risk
Power users consume 100× more than they pay.
Provider lock-in
Hard-coded SDKs make switching painful.
No runtime quotas
Nothing stops abuse between billing cycles.
No revenue sharing
Platforms can't take a margin on AI usage.
Billing complexity
Metering, top-ups and invoices become a roadmap.
Relay turns AI usage into infrastructure
Wallets, quotas, routing and fees behind a single OpenAI-compatible endpoint.
AI Gateway
One endpoint in front of every LLM provider.
Wallets & Top-Ups
Prepaid balances per tenant, user or workspace.
Runtime Quotas
Throttle by tokens, requests, cost or model class.
Provider Routing
Failover, fallbacks and routing rules at the edge.
Platform Fees
Add a margin or revenue share on every call.
OpenAI-Compatible API
Drop-in replacement — no SDK changes.
Usage Metering
Per-request cost attribution, exportable anywhere.
Multi-Tenant Keys
Issue scoped runtime keys per customer.
Two ways to monetize
Run Relay as a quota engine behind your existing billing, or hand it the wallet entirely.
Your app handles billing.
Relay enforces compute limits.
Plug Relay between your app and the LLMs. You keep Stripe, plans and seats — Relay enforces per-tenant token budgets and routes provider traffic.
- SaaS AI agents
- Internal copilots
- Enterprise assistants
Relay handles wallets,
top-ups and AI usage billing.
Full prepaid pipeline: hosted top-ups, balance enforcement, invoicing and fee splits. You ship product — Relay ships the cash register.
- Shared AI assistants
- AI developer platforms
- Embedded AI products
Monetize AI usage without building billing infrastructure
Take a margin on every top-up — Relay handles the splits, refunds and accounting.
A drop-in OpenAI-compatible endpoint
Change the baseURL, issue a tenant key, ship. Quotas, routing and metering happen at the edge.
// 1. Point any OpenAI SDK at Relay
const client = new OpenAI({
baseURL: "https://api.relaygateway.ai/v1",
apiKey: "relay_tenant_xxx",
})
// 2. Relay enforces wallet + quotas, routes the call,
// bills the tenant and splits platform fees.
await client.chat.completions.create({
model: "gpt-5",
messages: [{ role: "user", content: "Hello" }],
})
/v1/chat/completions200 OK · 412ms · $0.0042Teams shipping AI as a product
Everything you need to ship
Add AI monetization infrastructure to your product.
Stop building billing pipelines. Ship AI features and let Relay handle the wallet, the quotas and the routing.