Prompt Engineering Guide

Free reference site covering prompting techniques, papers and model notes

AI Prompt ToolsFree planOverseasβ˜…β˜…β˜…β˜†β˜† 3.0

What is Prompt Engineering Guide?

The Prompt Engineering Guide, maintained by DAIR.AI, is a free reference site that collects prompting techniques, research papers and model-specific guidance in one place. It explains ideas such as chain-of-thought, self-consistency and retrieval augmentation with worked examples, and links out to the papers behind them, which makes it useful for both quick lookups and deeper reading. Content covers text models as well as image and evaluation topics. It is documentation rather than software: there is no account and no evaluation harness, so it complements testing tools rather than replacing them.

Last updated: 2026-09-20. This site only provides an index; for exact features, pricing, and licensing, see the official website.

Key features

  • Technique reference covering chain-of-thought and related methods
  • Worked examples for each technique with input and output shown
  • Paper links and citations behind the main techniques
  • Sections on image prompting, evaluation and safety topics
  • Model-specific notes for common commercial and open models
  • Free to read with no account, paywall or tracking wall

Pros & cons

Strengths

  • Free and vendor-neutral reference material
  • Connects techniques to the underlying research papers
  • Good shared vocabulary for a team discussing prompt quality

Watch out for

  • Documentation, not a tool: nothing runs or evaluates prompts
  • Updates lag behind the newest model releases
  • Coverage of some topics is broad but not deep

Best for & use cases

learning prompt techniques, team onboarding, research pointers and teaching

If you're comparing similar products, check the alternatives below, or browse all tools in the AI Prompt Tools category.

FAQ

Is the guide tied to one model provider?

No. It covers techniques and models from several providers, which makes it a reasonable shared reference. Specific examples can still drift out of date as model behaviour changes.

Does it include exercises or code?

Some sections include runnable notebooks and examples, particularly for retrieval and evaluation topics. The site is primarily explanatory, so expect to bring your own environment to run anything.

How current is the material?

It is actively maintained but cannot track every model release. Treat newer model features as something to verify against official documentation rather than assuming the guide covers them.