⚡ L402 Payment Infrastructure

The Payment Layer for AI

Pay-per-request AI access via Lightning Network. No accounts. No API keys. No credit cards. Built for developers and autonomous agents.

⚡ Get Started View Documentation →

Why LightningProx

90 seconds to your first AI response

No email. No phone number. No credit card. No API key. Pay with Lightning, get instant AI access.

Traditional API Setup

10–15 min

LightningProx

curl → pay invoice → get response

90 sec

Developer Experience

Three commands. That's it.

Top up with Lightning, create a spend token, and start making AI requests. No SDK required.

        
        
        
        terminal
      

# 1. Top up — get a Lightning invoice
curl -X POST https://lightningprox.com/v1/topup \
  -d '{"amount_sats": 500}'

# 2. Pay the invoice, then create a token
curl -X POST https://lightningprox.com/v1/tokens \
  -d '{"charge_id": "YOUR_CHARGE_ID"}'

# 3. Use it — instant AI
curl -X POST https://lightningprox.com/v1/messages \
  -H "X-Spend-Token: lnpx_your_token" \
  -d '{"model": "claude-sonnet-4-20250514",
       "max_tokens": 100,
       "messages": [{"role": "user",
         "content": "Hello!"}]}'
      

⚡

Pay Per Request

No subscriptions. No minimums. Each request generates a Lightning invoice. Pay exactly what you use.

🎫

Spend Tokens

Pay once, use many times. Prepaid tokens for frictionless multi-request sessions. Perfect for agents.

🔌

MCP Server

Native MCP integration. Give Claude Desktop, Cursor, or any AI agent a Lightning wallet.

Infrastructure

Built for the agent economy

When millions of AI agents need to make billions of micro-requests, credit cards and API keys don't scale. Lightning does.

⚡

L402 Protocol

HTTP 402 Payment Required. Request, invoice, pay, response. The web's native payment flow.

🔒

Zero Auth

Payment is authentication. No accounts, no signups, no stored credentials. Truly permissionless.

🚀

Streaming

Real-time token-by-token responses via SSE. Feels 5x faster than waiting for a complete response.

💾

Smart Caching

50% discount on cached responses. Same query, half the cost. Automatic and transparent.

🤖

Multi-Model

Claude, GPT-4o, and more through a single endpoint. Switch models with one parameter.

🌐

Agent-Native

Designed for autonomous AI agents. No human-in-the-loop required for payment or auth.