Skip to content

providers.mdx

Latest commit

 

History

History
182 lines (132 loc) · 3.81 KB

providers.mdx

File metadata and controls

182 lines (132 loc) · 3.81 KB
Raw
title Setting Up as a Provider
description How to set up and configure the Routstr proxy for model providers

Setting Up as a Provider

This guide outlines how to set up the Routstr proxy to offer AI models through the Routstr network.

Overview

The Routstr proxy is a reverse proxy that sits in front of any OpenAI-compatible API endpoint to handle API key payments using Cashu tokens or Lightning invoices. As a provider, you can use this to monetize access to your AI models without setting up complex payment infrastructure.

Installation

Quick Start with Docker

The fastest way to get started is using Docker:

docker run -p 8000:8000 ghcr.io/routstr/proxy

Environment Configuration

The proxy requires configuration through environment variables:

# Model API endpoint configuration
MODEL_ENDPOINT_URL=https://api.openai.com/v1  # Or your self-hosted model API
OPENAI_API_KEY=your_api_key                   # If your endpoint requires authentication

# Payment settings
CASHU_MINT_URL=https://your-mint.com
LIGHTNING_ADDRESS=your@lightningaddress.com

# Network settings
PORT=8000                                     # Port to listen on

You can provide these in a .env file and use:

docker run -p 8000:8000 --env-file .env ghcr.io/routstr/proxy

Using docker-compose

For production deployments, you can use docker-compose:

version: '3'
services:
  routstr-proxy:
    image: ghcr.io/routstr/proxy
    ports:
      - "8000:8000"
    volumes:
      - ./models.json:/app/models.json
    environment:
      - MODEL_ENDPOINT_URL=https://api.openai.com/v1
      - OPENAI_API_KEY=your_api_key
      - CASHU_MINT_URL=https://your-mint.com
      - LIGHTNING_ADDRESS=your@lightningaddress.com
    restart: unless-stopped

Model Configuration

Configure your models in a models.json file:

[
  {
    "id": "gpt-4",
    "pricing": {
      "prompt": 0.03,
      "completion": 0.06
    },
    "contextWindow": 8192
  },
  {
    "id": "claude-3-opus",
    "pricing": {
      "prompt": 0.015,
      "completion": 0.075
    },
    "contextWindow": 200000
  }
]

Mount this file when running the Docker container:

docker run -p 8000:8000 -v $(pwd)/models.json:/app/models.json ghcr.io/routstr/proxy

Payment Integration

Cashu Integration

The proxy can integrate with any Cashu mint. Set the mint URL in your environment variables:

CASHU_MINT_URL=https://8333.space:3338

Lightning Integration

To accept Lightning payments, set your Lightning address:

LIGHTNING_ADDRESS=you@lightningaddress.com

Advanced Configuration

Custom Pricing

You can set different pricing for different models:

{
  "id": "gpt-4-turbo",
  "pricing": {
    "prompt": 0.01,
    "completion": 0.03
  }
}

TLS Configuration

For production deployments, configure TLS:

# Environment variables for TLS
TLS_CERT_FILE=/path/to/cert.pem
TLS_KEY_FILE=/path/to/key.pem

Proxy & Networking

For privacy-enhanced deployments, you can route through SOCKS5 or Tor:

# Route through SOCKS5
SOCKS5_PROXY=127.0.0.1:9050

Admin Dashboard

The proxy includes an admin dashboard accessible at:

http://your-server:8000/admin

This dashboard provides:

  • Current balance and transaction history
  • Request metrics and traffic analysis
  • Model usage statistics
  • Configuration settings

Nostr Integration

The proxy can announce your service on the Nostr network:

# Nostr announcement settings
NOSTR_PRIVATE_KEY=your_nsec_or_hex_key
NOSTR_RELAYS=wss://relay1.com,wss://relay2.com
NOSTR_ANNOUNCEMENT_INTERVAL=3600  # Announce every hour

Next Steps