The Prompt Engineering Guide is a comprehensive, open-source educational resource dedicated to the science and practice of designing effective prompts for large language models (LLMs). Maintained by DAIR.AI, it has grown into one of the most widely referenced references in the field, having reached over 3 million learners worldwide. The guide combines structured lessons, research summaries, and hands-on notebooks to help readers understand how to communicate with models like GPT-4, Llama, and Gemini more effectively.
What it covers
Content ranges from foundational concepts to advanced techniques, including Zero-Shot and Few-Shot prompting, Chain-of-Thought reasoning, Retrieval-Augmented Generation (RAG), ReAct, and other agentic patterns. It also addresses practical applications such as function calling, code generation, and synthetic data creation, as well as important topics like adversarial prompting and the risks and limitations of prompting-based approaches. Model-specific guides help readers apply these techniques across different LLM providers.
Who it's for
This resource suits a broad global audience: students and hobbyists just starting to explore LLMs, developers building production applications on top of language models, and researchers who want a structured overview of prompting techniques and the papers behind them. The material is organized progressively, so beginners can start with the fundamentals while more experienced readers can jump directly to advanced sections.
The project is released under an open license and is available as both a GitHub repository and a companion website, with translations into more than a dozen languages to support learners internationally. A companion lecture (with video, slides, and executable notebooks) and self-paced courses are also available for those who want a more guided learning path.