WebLLM is a high-performance in-browser language model inference engine that utilizes WebGPU for hardware acceleration, enabling powerful language model operations to be executed directly in web browsers without server-side processing. This project aims to seamlessly integrate large language models (LLMs) into the client side, resulting in cost reduction, enhanced personalization, and privacy protection. It supports various models, is compatible with the OpenAI API, is easy to integrate into projects, and supports real-time interaction and streaming, making it an ideal choice for building personalized AI assistants.