This project is an open-source Python library by the PyTorch team, designed to streamline the operation of Large Language Models (LLMs), enabling easy running of LLMs chat applications on platforms like desktops, servers, iOS, and Android. Users can interact with large models directly through the command line or initiate a WebUI for browser-based usage. Additionally, based on the PTE files generated by ExecuTorch, it can generate LLMs chat applications suitable for mobile devices, supporting large models such as Llama 3 and Mistral.