Configuration

Environment Variables

Variable	Description	Default
PORT	Server port	8082
NODE_ENV	Environment mode	development
OPENAI_API_KEY	OpenAI API key	-
OPENAI_DEFAULT_MODEL	Default OpenAI model	gpt-5.2
RATE_LIMIT_ENABLED	Enable rate limiting	false
RATE_LIMIT_TYPE	Rate limiter type (memory or redis)	memory
REDIS_URL	Redis connection URL	-
CORS_ORIGIN	CORS allowed origins	*
ROUTE_PREFIX	Route prefix for all endpoints	/ai
CONFIG_FILE	Path to JSON config file	-

Standalone Server

The package ships a ready-to-run entry point at jodit-ai-adapter/run. This is the fastest way to spin up a local server — no custom code required:

OPENAI_API_KEY=sk-... node -e "import('jodit-ai-adapter/run')"

It reads all settings from environment variables (see table above). To use a JSON config file instead, set CONFIG_FILE:

CONFIG_FILE=./config.json node -e "import('jodit-ai-adapter/run')"

Configuration File Format

{
  "port": 8082,
  "routePrefix": "/ai",
  "providers": {
    "openai": {
      "type": "openai",
      "apiKey": "sk-...",
      "defaultModel": "gpt-5.2"
    }
  },
  "rateLimit": {
    "enabled": true,
    "type": "memory",
    "maxRequests": 100,
    "windowMs": 60000
  }
}

Docker

The Docker image uses the same standalone entry point internally:

docker run -p 8082:8082 -e OPENAI_API_KEY=sk-... xdsoft/jodit-ai-adapter

With a config file:

docker run -p 8082:8082 \
  -v ./config.json:/app/config.json \
  -e CONFIG_FILE=/app/config.json \
  xdsoft/jodit-ai-adapter

CORS Configuration

By default the server allows all origins (*). For fine-grained control, use the cors object (it takes precedence over the legacy corsOrigin / CORS_ORIGIN setting):

Option	Type	Default	Description
origin	string \| string[] \| RegExp	*	Allowed origin(s)
methods	string	GET, POST, OPTIONS, PUT, DELETE	Allowed HTTP methods
allowedHeaders	string	Content-Type, Authorization, x-api-key, x-requested-with	Allowed request headers
credentials	boolean	true	Send Access-Control-Allow-Credentials
maxAge	number	86400	Preflight cache duration in seconds

JSON config file

{
  "cors": {
    "origin": "https://app.example.com",
    "methods": "GET, POST",
    "credentials": true,
    "maxAge": 3600
  }
}

Programmatic usage

import { createApp } from 'jodit-ai-adapter';

const { app } = createApp({
  cors: {
    origin: ['https://app.example.com', 'https://admin.example.com'],
    methods: 'GET, POST',
    allowedHeaders: 'Content-Type, Authorization',
    credentials: true,
    maxAge: 3600
  },
  providers: { /* ... */ }
});

Tip

For simple cases you can still use just corsOrigin: '*' or CORS_ORIGIN=https://example.com. The cors object is only needed when you want to customise methods, headers, credentials or max-age.

Programmatic Configuration

When using the package as a library in your own app:

import { start } from 'jodit-ai-adapter';

const { app, cleanup } = await start({
  port: 8082,
  config: {
    routePrefix: '/api/v1',
    providers: {
      openai: {
        type: 'openai',
        apiKey: process.env.OPENAI_API_KEY,
        defaultModel: 'gpt-5.2'
      }
    },
    rateLimit: {
      enabled: true,
      type: 'memory',
      maxRequests: 100,
      windowMs: 60000
    }
  }
});

// To gracefully shut down (stops server + releases resources):
// await cleanup();

See Rate Limiting for advanced rate limiting options (Redis, per-user limits, etc.).