LIDA is an open-source tool developed by Microsoft Research for the automatic generation of grammar-agnostic visualizations and infographics from datasets using large language models (LLMs) and image generation models (IGMs). It features a conversational interface and modular workflow—summarizing data, exploring visualization goals, generating visualizations in various grammars (e.g., Python, R, C++), and creating stylized infographics. LIDA is intended for data scientists, analysts, educators, and anyone needing automated and explainable data visualizations and exploration, and is available via Python API.
Visit LIDA's official website for product details and getting started.
Access the source code, report issues, and contribute to LIDA on GitHub.