Google (Gemini)
Gemini integration via the Generative Language API with automatic format conversion.
Configuration
1export GEMINI_API_KEY=AI...Supported Models
| Model | Input (per 1M) | Output (per 1M) |
|---|---|---|
| gemini-2.5-pro | $1.25 | $10.00 |
| gemini-1.5-pro | $1.25 | $5.00 |
| gemini-2.0-flash | $0.10 | $0.40 |
| gemini-1.5-flash | $0.075 | $0.30 |
Format Conversion
RelayPlane automatically converts between OpenAI chat format and Gemini's native format:
- OpenAI
messages→ Geminicontents(role mapping: assistant → model) - OpenAI
systemmessage → GeminisystemInstruction - OpenAI
tools→ GeminifunctionDeclarations - Gemini
functionCallresponses → OpenAItool_calls - Multimodal content (images) with base64 inline data support
Endpoint
Base URL: https://generativelanguage.googleapis.com/v1beta
Authentication uses the API key as a query parameter (?key=...).
Model Detection
Models starting with gemini- or palm- are automatically routed to Google.
Both streaming (
streamGenerateContent) and non-streaming (generateContent) are supported with full SSE-to-OpenAI format conversion.