​
Getting Started as a Provider

This guide will help you set up as a model provider on Routstr. By offering your AI models through the Routstr network, you can reach users worldwide and receive micropayments without complex payment infrastructure.

​
What You Can Offer

As a provider, you can offer:

  • Access to any OpenAI-compatible API endpoint
  • Open-source models you’re hosting (Llama, Mistral, etc.)
  • Proprietary models with custom pricing
  • Specialized models with unique capabilities

​
Benefits for Providers

Joining Routstr as a provider gives you:

  • Instant Monetization: Receive Lightning or Cashu micropayments without setting up complex payment infrastructure
  • Global Reach: Connect with users worldwide through the decentralized marketplace
  • Zero Overhead: No need for user management, subscriptions, or billing systems
  • Full Control: Set your own pricing, capacity limits, and access policies

​
Setting Up Your Provider Node

​
Prerequisites

  • An OpenAI-compatible API endpoint (e.g., your locally hosted model, a proxied commercial model, etc.)
  • Docker (recommended) or a Linux server
  • A Lightning address or Cashu mint to receive payments

​
Quick Setup with Docker

The simplest way to set up a Routstr provider is using the official Docker image:

docker run -p 8080:8080 ghcr.io/routstr/proxy

This will start a proxy server that can accept Cashu token payments and forward requests to your OpenAI-compatible API endpoint.

​
Configuration

The proxy requires configuration to connect to your model endpoint and set up payments. You can create a .env file with the following settings:

# API endpoint settings
OPENAI_API_KEY=your_model_api_key
MODEL_ENDPOINT_URL=http://your-model-endpoint:8000/v1

# Payment settings
CASHU_MINT_URL=https://your-cashu-mint.com
LIGHTNING_ADDRESS=your@lightning.address

Then run the Docker container with your configuration:

docker run -p 8080:8080 \
  --env-file .env \
  ghcr.io/routstr/proxy

​
Manual Setup

If you prefer to set up the provider manually:

  1. Clone the Routstr proxy repository:

    git clone https://github.com/Routstr/proxy.git
cd proxy

  2. Create a .env file with your configuration (see example above)

  3. Start the proxy using docker-compose:

    docker-compose up

​
Configuring Your Models

You can customize the pricing and configuration for your models by editing the models.json file:

[
  {
    "id": "gpt-4",
    "pricing": {
      "prompt": 0.03,
      "completion": 0.06
    }
  },
  {
    "id": "llama-3-70b-chat",
    "pricing": {
      "prompt": 0.0007,
      "completion": 0.0014
    }
  }
]

This configuration specifies the pricing per 1K tokens for input (prompt) and output (completion).

​
Monitoring Your Provider

You can monitor your provider using the built-in dashboard:

# Access the dashboard
open http://localhost:8080/admin

The dashboard provides information about:

  • Request volume and traffic
  • Revenue generated
  • Model usage statistics
  • Payment status

​
Announcing Your Provider

After your provider is running, you can announce it to the Routstr network via Nostr:

  1. Configure your provider with your Nostr private key or npub
  2. Enable automatic announcements in the configuration
  3. Set a custom name and description for your provider

​
Next Steps