Technical

What is Zero-Shot Prompting?

Asking an AI model to complete a task with no examples of the desired output.

Definition

Zero-shot prompting is when you give an AI model a task description with no examples of the expected output — you're relying on the model's pre-trained knowledge to infer what you want. It's the simplest prompting approach. Models have become increasingly capable zero-shot; many tasks that previously required few-shot examples now work zero-shot with modern LLMs.

Example

Zero-shot: 'Classify this email as spam or not-spam: [email text].' The model has no example of what spam vs not-spam looks like in your format — it uses its training to infer.

Zero-Shot Prompting vs few-shot: What's the difference?

Zero-Shot Prompting

Asking an AI model to complete a task with no examples of the desired output.

few-shot

Zero-shot: no examples provided. Few-shot: a small number of input-output examples are included in the prompt to guide the model.

Related terms

Back to glossary