gemini-webapi - PyPI

Introduction

gemini-webapi is a Python-based asynchronous wrapper library designed to give developers a clean, efficient way to interact with Google's Gemini large language model. With this library, you can easily call Gemini's native Web API and quickly build applications for intelligent conversation, content generation, code assistance, and more. The library is built with async support at its core, ensuring strong performance even in high-concurrency or I/O-intensive workloads.

Key Features

• Async call support: Built on asyncio and aiohttp, it fits seamlessly into the modern Python async ecosystem.
• Multimodal interaction: Supports text, image, audio, and other input formats, fully leveraging Gemini's multimodal understanding capabilities.
• Streaming responses: Built-in streaming output lets you receive generated results in real time, improving the user experience.
• Conversation management: Provides simple session context management for easy multi-turn dialogue and memory features.
• Error retry and timeout control: Includes an intelligent retry mechanism and configurable timeout policies to boost call stability.
• Lightweight with minimal dependencies: Core dependencies are minimal, installation is simple, and you can get started with almost zero configuration.

Highlights

• The standout feature of gemini-webapi is its purely asynchronous architecture. Unlike traditional synchronous wrappers, this library embraces the async programming model from the ground up, so you can handle a large number of concurrent requests without worrying about thread blocking or wasted resources.
• The interface design is highly Pythonic, with intuitive function naming and thorough documentation, making it easy even for beginners to async programming to pick up quickly.
• Compared to the official SDK or other third-party wrappers, gemini-webapi significantly reduces integration complexity while maintaining full feature completeness.

Who It's For

Python developers who want to quickly integrate Gemini capabilities into existing async projects (such as FastAPI, Sanic, or Tornado); AI application builders working on chatbots, content generation tools, or intelligent assistants; researchers and educators who need to batch test or analyze Gemini model outputs, as well as instructors teaching the combination of async programming and AI; and tech enthusiasts interested in Google's Gemini model who want to experiment with Python-based development.