Last updated: April 5, 2026 · Prompting & Usage · by Daniel Ashford

What is Prompt?

QUICK ANSWER

The text input you send to an LLM to get a response.

Definition

A prompt is the text input provided to a language model to elicit a response. Prompts can range from simple questions to complex multi-part instructions with examples, constraints, and formatting requirements. The quality of the prompt significantly influences the quality of the output.

How It Works

Prompts typically consist of a system prompt (persistent behavior instructions), user messages (the actual task), and optionally few-shot examples (demonstrations of desired output). Effective prompting is both an art and a science — techniques like chain-of-thought prompting, role-playing, and structured output formatting can dramatically improve results.

Example

Simple prompt: "What is the capital of France?" Complex prompt: "You are a financial analyst. Analyze the following earnings report, focusing on revenue growth, margin trends, and forward guidance. Present in bullet points with specific numbers."

Related Terms

System Prompt
Persistent instructions that define how the model should behave.
Few-Shot Prompting
Providing examples of desired input-output pairs in the prompt.
Chain-of-Thought (CoT)
A prompting technique that asks the model to show its reasoning step by step.
Prompt Engineering
The skill of crafting effective prompts to get the best results from an LLM.
Temperature
A setting that controls how creative or deterministic responses are.

See How Models Compare

Understanding prompt is important when choosing the right AI model. See how 12 models compare on our leaderboard.

View Leaderboard →Our Methodology
← Browse all 47 glossary terms
DA
Daniel Ashford
Founder & Lead Evaluator · 200+ models evaluated