AgentSDK Gemini Adapter

A Python adapter for using Google Gemini with OpenAI Agents SDK. This package provides a seamless integration between Google's Gemini models and the OpenAI Agents SDK.

Installation

pip install agentsdk-gemini-adapter

Quick Start

1. Set up your Gemini API key in a .env file:

GEMINI_API_KEY=your_gemini_api_key_here

2. Import and use the config:

from agentsdk_gemini_adapter import config

# Now you can use the config with OpenAI Agents SDK
# The config includes a pre-configured Gemini model

Features

• Pre-configured Gemini 2.0 Flash model
• Seamless integration with OpenAI Agents SDK

Requirements

• Python 3.7+
• Gemini API Key
• OpenAI Agents SDK

Dependencies

This package requires openai-agents to be installed separately:

pip install openai-agents

Note: python-dotenv will be installed automatically as a dependency of openai-agents.

Usage Example

from agents import Agent, Runner
from agentsdk_gemini_adapter import config

# Create an agent using OpenAI Agents SDK
agent = Agent(
    name="Assistant",
    instructions="You are a helpful assistant.",
)

# Pass the Gemini configuration in run_config to any Runner method
result = Runner.run_sync(agent, "Write a haiku about recursion in programming.", run_config=config)
print(result.final_output)

Configuration

The package automatically:

• Loads your GEMINI_API_KEY from environment variables
• Configures the Gemini client with the correct base URL
• Sets up the Gemini 2.0 Flash model
• Creates a ready-to-use RunConfig object

License

MIT License - see LICENSE file for details.

Contribution

Contributions are welcome! Please feel free to submit a Pull Request.