Gateway Server

Deploy the Mako Gateway on your own infrastructure.

Prerequisites

• Node.js 18+
• An Ollama instance (for the 8B model) or RunPod endpoint (for 32B)

Setup

1. Clone the repository

   git clone https://github.com/DeepMako/mako.git
   cd mako/mako-platform/gateway
   
2. Install dependencies

   npm install
   
3. Configure environment

   cp .env.example .env
   

   Edit .env with your configuration. See Environment Variables for the full reference.

4. Start the server

   npm start
   

   You should see:

   🦈 Mako Gateway running on http://localhost:3000
   📡 Model: mako @ http://localhost:11434/api/chat
   🆓 Free mode: ON — open access, no auth required
   

API routes

The gateway exposes these endpoints:

Method	Path	Description
POST	/v1/chat/completions	Chat completion (OpenAI-compatible)
GET	/v1/models	List models
GET	/v1/tools	List available tools
POST	/v1/tools/execute	Execute a tool directly
GET	/credits/balance	Check credit balance
POST	/credits/grant	Grant credits (dev only)
POST	/auth/api-keys	Create an API key
GET	/health	Health check

Database

The gateway uses SQLite by default, stored at ./data/mako.db. The database manages:

• User accounts (keyed by wallet address)
• API keys (hashed)
• Credit balances and usage logs

For production, set DATABASE_PATH to a persistent volume.

Deployment

Railway

The repository includes railway.json and nixpacks.toml for one-click Railway deployment:

cd mako/mako-platform/gateway
railway up

Docker

Build and run with any container platform:

docker build -t mako-gateway .
docker run -p 3000:3000 --env-file .env mako-gateway

Other platforms

The gateway is a standard Express.js app. It works on Render, Heroku, Fly.io, or any Node.js hosting. Just set the environment variables and run npm start.