• 中文
• video

This is a proxy server designed specifically for the Google Gemini API. It allows you to securely consolidate multiple API keys into a single endpoint and randomly select one for use with each request. This is useful for managing keys, load balancing, and integrating with front-end applications.

• Multi-Key Management: Pass multiple Google AI API keys, separated by commas, in the x-goog-api-key header.
• Random Key Selection: A key is randomly selected from your provided list for each request, helping to distribute the load.
• Request Forwarding: Seamlessly forwards all requests to the Google Generative Language API (https://generativelanguage.googleapis.com).
• Flexible Deployment: Optimised for Vercel, but also supports deployment using Docker.

We highly recommend using Docker + Claw Cloud for a quick and easy one-click deployment.

1. Click the "Deploy to Vercel" button above.
2. Follow the instructions on Vercel to clone this repository and deploy it.
3. Once deployed, you will receive a dedicated proxy URL.

You can also use Docker to deploy on any supported platform, such as Claw Cloud.

docker run -d \
  -p 80:80 \
  --name gemini-proxy \
  --restart unless-stopped \
  ghcr.io/spectre-pro/gemini-proxy

3. Your proxy server will be running at http://localhost:80.