Few-Shot Prompting
Giving the model a handful of worked examples inside your prompt so it copies the pattern.
Often the fastest way to lock in a format or style without any fine-tuning.
Frequently asked questions
What is few-shot prompting?
Few-shot prompting means including a handful of worked examples right in your prompt so the model copies the pattern. If you show it two or three examples of the input and the output you want, it tends to follow that shape.
When should I add examples to my prompt?
Add examples when you need a specific format or style locked in, since showing is often faster and clearer than describing. It is frequently the quickest way to get consistent results without any fine-tuning.
How is few-shot different from zero-shot prompting?
Zero-shot means asking with no examples and relying on the model's instructions and training, while few-shot means giving it a few examples to imitate. Zero-shot is the everyday default; you switch to few-shot when the format or style needs to be exact.
New to all this? Start with what an AI agent really is, browse the full glossary, or explore the learning hub.
← Back to the glossary