🔓 Free ChatGPT API

A high-performance, OpenAI-compatible API server written in Go that gives you free access to ChatGPT — including GPT-5.5, GPT-4o, GPT Image Generation, and more. No API key required. Fully automated session authentication via a lightweight Chrome extension.

---

🤔 Why Use This?

Problem	Solution
OpenAI API costs $20+/month	This is 100% free
API keys get leaked and abused	No API keys needed
Rate limits on free tiers	Uses your own ChatGPT session
Complex auth setup	One-click Chrome extension auto-syncs cookies
Want OpenAI-compatible endpoints	Drop-in replacement for /v1/chat/completions

---

✨ Features

• 💬 OpenAI Chat Completions — Standard /v1/chat/completions endpoint, works with any OpenAI client library
• 🎨 Image Generation — Generate images via ChatGPT's built-in DALL-E, auto-downloaded to output/
• 🔄 Conversation Persistence — Same conversation continues across requests (no rate limit issues from new chats)
• 🔌 Chrome Cookie Bridge — Lightweight extension auto-syncs your ChatGPT session cookies via WebSocket
• ⚡ Zero Config — Just set port and model in .env, everything else is hardcoded for reliability
• 🧠 Multi-Model Support — GPT-5.5, GPT-4o, GPT-4o-mini, o3, o4-mini and more

---

🏗️ How It Works

┌─────────────┐     ┌──────────────────┐     ┌─────────────────┐
│  Your App    │────▶│  Go API Server   │────▶│  ChatGPT Web    │
│  (curl/SDK)  │◀────│  (localhost:9225) │◀────│  (chatgpt.com)  │
└─────────────┘     └──────────────────┘     └─────────────────┘
                           ▲
                           │ WebSocket (cookies)
                    ┌──────┴───────┐
                    │   Chrome     │
                    │  Extension   │
                    └──────────────┘

1. Chrome Extension captures your ChatGPT session cookies and sends them to the Go server via WebSocket
2. Go Server receives your API request at /v1/chat/completions
3. Server forwards the request to ChatGPT's web backend using your session cookies
4. Response is formatted as a standard OpenAI API response and returned to your app

---

🚀 Quick Start & Docker Deployment

1. Build and Run in Docker (Recommended)

You can run the API server inside a Docker container:

# Build the Docker image
docker build -t chatgpt-free-api:latest .

# Run the container (binds to host port 9225)
docker run -d \
  --name chatgpt-free-api \
  -p 9225:9225 \
  --restart unless-stopped \
  chatgpt-free-api:latest

To build/run locally without Docker:

go run .

2. Install Chrome Extension

1. Open chrome://extensions/ in Chrome.
2. Enable Developer mode (top right toggle).
3. Click Load unpacked (top left button) → select the gpt-extension/ folder in this repository.
4. Open chatgpt.com and log in.
5. Click on the extension icon. By default, it connects to port 9225.

   [!TIP] If you configured the server to run on a different port, you can change the Server Port inside the extension popup UI directly. It will dynamically save and reconnect to the new port without reloading the extension.

3. Start Using

# Chat (OpenAI-compatible)
curl -X POST http://localhost:9225/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

---

📡 API Reference

POST /v1/chat/completions (OpenAI-Compatible)

Request:

{
  "model": "gpt-5-5",
  "messages": [
    {"role": "user", "content": "Your prompt here"}
  ],
  "conversation_id": "optional-to-continue-chat"
}

Response:

{
  "id": "chatcmpl-abc123",
  "object": "chat.completion",
  "created": 1234567890,
  "model": "gpt-5-5",
  "conversation_id": "6a404fee-921c-83ee-850e-f2a582c56e7c",
  "choices": [
    {
      "index": 0,
      "message": {"role": "assistant", "content": "Hello!"},
      "finish_reason": "stop"
    }
  ]
}

POST /api/chat (Native Endpoint)

{
  "prompt": "Your message",
  "conversation_id": "",
  "model": "gpt-5-5"
}

Other Endpoints

Endpoint	Method	Description
/v1/chat/completions	POST	OpenAI-compatible chat
/api/chat	GET/POST	Native chat endpoint
/api/chat/bulk	POST	Batch multiple prompts
/api/chat/edit	POST	Edit last message
/api/sniffs	GET	View request logs
/health	GET	Server health check

---

⚙️ Configuration

Only 2 values in .env — everything else is hardcoded for stability:

LISTEN_ADDR=127.0.0.1:9225
DEFAULT_MODEL=gpt-5-5

Variable	Default	Description
LISTEN_ADDR	127.0.0.1:9225	Server address and port
DEFAULT_MODEL	gpt-5-5	Default ChatGPT model

---

📂 Project Structure

chatgpt-free-api/
├── main.go                 # Entry point
├── .env                    # Server config (port + model)
├── chatgpt/                # Core Go source code
│   ├── agent.go            # Server bootstrap & routing
│   ├── handlers.go         # HTTP request handlers
│   ├── chatgpt.go          # ChatGPT web client & WebSocket bridge
│   ├── config.go           # .env config loader
│   ├── cookies.go          # Cookie management
│   ├── api_bridge.go       # Image download bridge
│   ├── extension.go        # Chrome extension WebSocket handler
│   ├── helpers.go          # Utility functions
│   └── sniff.go            # Request logging/sniffing
├── gpt-extension/          # Chrome extension
│   ├── manifest.json
│   ├── background.js
│   ├── popup.html
│   └── popup.js
└── output/                 # Auto-downloaded generated images

---

🔒 Security Note

This tool uses your own ChatGPT session — it does not share credentials with any third party. All communication happens between your local machine and ChatGPT's servers. The Chrome extension only sends cookies to localhost.

---

📜 License

MIT License — free to use, modify, and distribute.