Convert Google AI Studio web interface to an OpenAI-compatible API proxy server. Provides stable API access through Camoufox (anti-fingerprinting Firefox) and Playwright automation. Enhanced with a Python client library (gemini_client.py) for seamless integration and automated workflows.

This is an English translation and adaptation of the AIstudioProxyAPI project. Special thanks to the original author for creating this excellent proxy solution.

• OpenAI Compatible API: Fully compatible with /v1/chat/completions endpoint and mainstream clients
• Three-layer Streaming Response Mechanism: Integrated streaming proxy → External Helper → Playwright page interaction
• Intelligent Model Switching: Dynamically switch AI Studio models with complete parameter control
• Anti-fingerprinting Detection: Camoufox browser reduces detection risk
• Modern Web UI: Built-in testing interface, status monitoring, API key management
• Graphical Interface Launcher: GUI launcher simplifies configuration and process management
• Script Injection v3.0: Playwright native network interception, supports dynamic mounting of userscripts

• gemini_client.py: A Python client library for easy integration and cross-model validation
• gemeni_reviewer.py: Code review automation script using the proxy API
• AGENTS.md: AI system prompts and behavior guidelines - feel free to enhance and improve them

# 1️⃣ Clone and install
git clone https://github.com/yaron4u/AIstudioProxyAPIClient.git
cd AIstudioProxyAPI
poetry install

# 2️⃣ Configure environment
cp AIstudioProxyAPI/.env.example .env
# Edit .env file with your configuration (optional)

# 3️⃣ First authentication and startup
poetry run python AIstudioProxyAPI/launch_camoufox.py --debug  # First authentication (requires Google login)
# After successful authentication, Ctrl+C to stop, then:
poetry run python AIstudioProxyAPI/launch_camoufox.py --headless

# Health check
curl http://127.0.0.1:2048/health

# Get model list
curl http://127.0.0.1:2048/v1/models

# Test chat
curl -X POST http://127.0.0.1:2048/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-2.5-pro-preview","messages":[{"role":"user","content":"Hello"}]}'

Access http://127.0.0.1:2048/ to use the built-in Web UI.

graph TD
    subgraph "Client Side"
        User["User"]
        WebUI["Web UI"]
        API_Client["API Client"]
    end

    subgraph "Launch & Configuration"
        GUI_Launch["gui_launcher.py"]
        CLI_Launch["launch_camoufox.py"]
        EnvConfig[".env Configuration"]
    end

    subgraph "Core Application"
        FastAPI_App["FastAPI Application<br/>api_utils/"]
        PageController["Page Controller<br/>browser_utils/"]
        StreamProxy["Stream Proxy<br/>stream/"]
    end

    subgraph "External Dependencies"
        CamoufoxInstance["Camoufox Browser"]
        AI_Studio["Google AI Studio"]
    end

    User --> GUI_Launch & CLI_Launch
    API_Client & WebUI --> FastAPI_App
    FastAPI_App --> PageController & StreamProxy
    PageController --> CamoufoxInstance --> AI_Studio
    StreamProxy --> AI_Studio

The project uses .env file for unified configuration management:

cp AIstudioProxyAPI/.env.example .env
# Edit .env file with your configuration

Detailed Configuration: See Environment Variables Reference

cd AIstudioProxyAPI/docker
cp .env.docker .env
# Edit .env file

docker compose up -d
docker compose logs -f

# Version update
bash update.sh

Detailed Guide: Docker Deployment Guide

• Quick Start Guide - 15-minute quick deployment 🎯
• Installation Guide - Detailed installation steps
• Authentication Setup Guide - First authentication setup
• Daily Usage Guide - Daily usage

• API Usage Guide - API endpoints and configuration
• OpenAI Compatibility Notes - Differences with OpenAI API 🔄
• Client Integration Examples - Code examples 💻
• Web UI Usage Guide - Web interface features
• Script Injection Guide - Userscript functionality (v3.0)

• Environment Variables Configuration Guide - Configuration management ⭐
• Complete Environment Variables Reference - All configuration items 📋
• Streaming Processing Modes Explained - Three-layer response mechanism
• Advanced Configuration Guide - Advanced features
• Troubleshooting Guide - Problem solving

• Project Architecture Guide - Modular architecture
• Developer Guide - Poetry, Pyright workflow

Using Open WebUI as an example:

1. Go to "Settings" → "Connections"
2. Add model, API Base URL: http://127.0.0.1:2048/v1
3. API Key: Leave empty or any character
4. Save and start chatting

from gemini_client import GeminiClient

client = GeminiClient()
response = client.ask("Explain quantum computing in simple terms")
print(response)

python gemeni_reviewer.py  # Reviews files listed in the script

You're welcome to create pull requests and contribute to this project! Whether it's:

• Bug fixes
• Feature enhancements
• Documentation improvements
• Additional client libraries
• Cross-model validation tools

All contributions are appreciated. Please ensure your code follows the project's standards and includes appropriate tests.

• Original project: CJackHwang/AIstudioProxyAPI

See the LICENSE file in the aistudioproxy/AIstudioProxyAPI directory.