Base URL: http://localhost:3131/v1
Local Proxy
0
Total Requests
Streamed
0
Processed Tokens
Daily Limit
Unlimited
Via ChatGPT Engine
Ready
Port 3131
Ready for connections
Request Activity
Real-time throughput metrics over session timeline
T-60m T-45m T-30m T-15m Now
1-Click Cursor Setup
Add custom OpenAI-compatible model in Cursor settings
Cursor → Settings → Models 
{
  "apiBase": "http://localhost:3131/v1",
  "apiKey": "free",
  "model": "gpt-5.6-luna" // or "gpt-5.6-luna-thinking"
}
Supports Cursor, Continue, Roo, Cline, & Python SDK
Recent Requests Stream
Real-time incoming chat completions via /v1/chat/completions
Time Prompt Preview Tokens (est) Status Action
No completions recorded yet. Send a prompt from Cursor or the Playground!
GPT-5.6 Luna Studio
Send a live prompt to test your local proxy. Toggle Deep Reasoning to trigger high-depth analytical problem solving.
Real-time SSE streaming via POST /v1/chat/completions Enter to send · Shift+Enter for newline
Account Authorization
Connect your ChatGPT session to enable the proxy bridge
Paste your entire DevTools Cookie header here to auto-populate chunks!
— OR ENTER CHUNKS SEPARATELY —
Quick Cookie Extraction Guide
Takes 15 seconds in Google Chrome or Edge
  1. Open chatgpt.com and sign in.
  2. Press F12 (or right click → Inspect) to open DevTools.
  3. Click the Application tab at the top.
  4. In the left sidebar, expand Cookies → click on https://chatgpt.com.
  5. Locate the cookies:
    __Secure-next-auth.session-token.0
    __Secure-next-auth.session-token.1
  6. Copy both values and paste them into the input fields on the left.
Privacy Guarantee: All tokens stay strictly in your local .browser_profile directory on this computer. Zero analytics or credentials leave your device.
Developer Integrations & Guides
Connect Cursor, Continue.dev, VS Code, Python SDK, or cURL
Cursor → Settings (Ctrl + ,) → Models → Custom OpenAI-Compatible 
// Cursor Settings -> Models -> Add OpenAI-Compatible Model
Base URL: http://localhost:3131/v1
API Key:  free

Models to add:
- gpt-5.6-luna
- gpt-5.6-luna-thinking
Tip: Set gpt-5.6-luna-thinking as your primary Composer model in Cursor for deep agentic reasoning.
~/.continue/config.json 
{
  "models": [
    {
      "title": "GPT-5.6 Luna",
      "provider": "openai",
      "model": "gpt-5.6-luna",
      "apiBase": "http://localhost:3131/v1",
      "apiKey": "free"
    },
    {
      "title": "GPT-5.6 Luna (Thinking)",
      "provider": "openai",
      "model": "gpt-5.6-luna-thinking",
      "apiBase": "http://localhost:3131/v1",
      "apiKey": "free"
    }
  ]
}
openai Python Package 
from openai import OpenAI

client = OpenAI(
    base_url="http://localhost:3131/v1",
    api_key="free"
)

response = client.chat.completions.create(
    model="gpt-5.6-luna-thinking",
    messages=[
        {"role": "user", "content": "Explain asynchronous programming in Python"}
    ],
    stream=True
)

for chunk in response:
    content = chunk.choices[0].delta.content or ""
    print(content, end="", flush=True)
print()
openai npm module 
import OpenAI from 'openai';

const openai = new OpenAI({
  baseURL: 'http://localhost:3131/v1',
  apiKey: 'free',
});

const stream = await openai.chat.completions.create({
  model: 'gpt-5.6-luna',
  messages: [{ role: 'user', content: 'Hello from Node.js!' }],
  stream: true,
});

for await (const chunk of stream) {
  process.stdout.write(chunk.choices[0]?.delta?.content || '');
}
Terminal HTTP Request 
curl http://localhost:3131/v1/chat/completions \
  -H "Authorization: Bearer free" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.6-luna-thinking",
    "stream": true,
    "messages": [
      {"role": "user", "content": "Hello from cURL!"}
    ]
  }'
Cline / Roo-Code VS Code Extension 
Provider: OpenAI-Compatible
Base URL: http://localhost:3131/v1
API Key:  sk-luna-your-key-here
Model ID: gpt-5.6-luna-thinking
Generated
0
Total API Keys
Active
0
Enabled Keys
Total Calls
0
Authenticated Requests
Optional
Open
V1 Route Enforcement
Require API Keys for /v1 Routes
When enabled, clients (Cursor, Continue, Python, cURL) must provide an Authorization: Bearer sk-luna-... header.
Disabled
Active Secret Keys
Manage client authorization credentials
Name / Label Secret Key Created Last Used Calls Status Actions
No API keys generated yet. Click "Create New Key" above.
Live Request Inspector
Real-time audit log of all completions and streamed token counts
Time Prompt Payload Snippet Est. Tokens Status Action
No requests recorded yet. Send a test message from the Playground!
Local Health & Latency Test
Ping local endpoints to verify bridge availability
GET /api/status Not tested
GET /v1/models Not tested
OpenAI HTTP Specs
Supported REST routes for extensions and SDKs
POST /v1/chat/completions SSE Streaming & JSON
GET /v1/models Model Discovery
GET /api/status Live Bridge Status
GET /api/logs Audit Logs