<2-3 sentences summarising what the project addresses and solves, the target audience, and the benefit.> This project provides a Dockerized proxy for Google's AI Studio build, integrating a headless browser and WebSocket proxy. It aims to offer a stable connection to Gemini-like models by leveraging robust cookie management, enabling users to interact with AI services via a local endpoint that mimics the official Gemini API. It's suitable for developers and researchers needing a reliable local interface for AI model interactions, especially those requiring tool usage like Google Search or code execution.

How It Works

The core mechanism involves a headless browser managed via Docker Compose. It utilizes stable cookies exported from fingerprint browsers to maintain authenticated sessions with AI services. A WebSocket proxy routes requests to the headless browser, which then forwards them to the target AI model endpoint (e.g., Gemini). This setup allows local applications to interact with the AI model as if they were directly calling the generativelanguage.googleapis.com API, but through the local 127.0.0.1:5345 endpoint.

Quick Start & Requirements

• Installation: Clone the repository and run docker compose up -d in the project root.
• Prerequisites: Docker and Docker Compose installed. A Google API key (AUTH_API_KEY) is required for Gemini access.
• Configuration:
  1. Export stable cookies (e.g., from a fingerprint browser) into the camoufox-py/cookies/ directory as a JSON file.
  2. Rename config.yaml.example to config.yaml and update the cookie_file path. Optionally configure the target url and proxy settings.
  3. Set the AUTH_API_KEY in docker-compose.yml.
• Access: The service is available at http://127.0.0.1:5345.
• Documentation: API interface mirrors Gemini API documentation: https://generativelanguage.googleapis.com.

Highlighted Details

• Provides a local proxy endpoint (http://127.0.0.1:5345) compatible with the Gemini API.
• Supports native Google Search and code execution tools when interacting with the proxy.
• Employs stable cookie export methods using fingerprint browsers for reliable authentication.
• Includes logging and resource monitoring capabilities via Docker.

Maintenance & Community

No specific information regarding maintainers, community channels (like Discord/Slack), sponsorships, or roadmap was found in the provided text.

Licensing & Compatibility

The license type is not specified in the provided README content. Compatibility for commercial use or linking with closed-source projects is undetermined without a license.

Limitations & Caveats

The proxy is specifically for Gemini-like models and requires a valid AUTH_API_KEY. Performance issues (slow inference) are attributed to the upstream AI Studio service, not the proxy itself. It relies heavily on the correct configuration and stability of the exported cookies. No explicit license information is provided, which may pose risks for adoption. Setup requires familiarity with Docker and API key management.