Route Google Gemini through Stockyard

Why proxy Google Gemini?

Google Gemini uses its own API format. Stockyard translates Gemini requests to and from the OpenAI-compatible format, so you can use Gemini through the same endpoint and SDK as OpenAI and Anthropic.

Gemini models are competitively priced, especially Flash variants. Proxying through Stockyard lets you route cost-sensitive requests to Gemini while keeping quality-critical requests on Claude or GPT-4o, all through one endpoint.

Quick start

# Install Stockyard
curl -fsSL stockyard.dev/install.sh | sh

# Set your Google Gemini API key
export GEMINI_API_KEY=your-key-here

# Start the proxy
stockyard
# Provider: google-gemini (from GEMINI_API_KEY)
# Proxy listening on :4200

# Send a request through the proxy
curl http://localhost:4200/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{"model":"gemini-2.5-pro","messages":[{"role":"user","content":"hello"}]}'

Good to know

Gemini model names are auto-detected. Use the full model name (e.g., gemini-2.5-flash) in your request body.

Use Gemini Flash as your cost optimizer

Gemini 2.5 Flash is one of the cheapest capable models available. With model aliasing, you can route cost-sensitive tasks to Flash while keeping quality-critical requests on GPT-4o or Claude:

# Route cheap tasks to Gemini, expensive tasks to GPT-4o
curl -X PUT http://localhost:4200/api/proxy/aliases \
  -d '{"alias":"cheap","model":"gemini-2.5-flash"}'
curl -X PUT http://localhost:4200/api/proxy/aliases \
  -d '{"alias":"quality","model":"gpt-4o"}'

Your app sends to cheap or quality. Change the backing model anytime without redeploying.