What Is a Prompt in AI?

A prompt is the text instruction you give an AI tool like ChatGPT, Claude, or Gemini to tell it what you want. It can be a question, a command, or a detailed set of instructions — and the AI's answer is based entirely on it.

Updated June 27, 2026

If you've ever typed a message into ChatGPT, you've written a prompt. The prompt is your half of the conversation: it's how you communicate your goal to the AI. Everything the AI produces is a response to that input, which is why the wording of your prompt has such a big effect on what you get back.

How prompts work

AI assistants are built on large language models — systems trained on enormous amounts of text to predict what words should come next. When you send a prompt, the model reads it and generates a response one piece at a time, guided by the patterns it learned during training. It isn't looking anything up or "thinking" the way a person does; it's producing the most fitting continuation of your text.

That's the key insight: the AI responds to what you actually wrote, not what you meant. A clear, specific prompt points it in the right direction. A vague one leaves it guessing.

The main types of prompts

• Questions — "What's a good way to remove a coffee stain?" The simplest kind of prompt.
• Instructions — "Summarize this article in three bullet points." You tell the AI to perform a task.
• Role prompts — "Act as a travel agent and plan a 3-day trip to Rome." You give the AI a persona to shape its tone and expertise.
• Example-based prompts — You show the AI a sample of the style or format you want, then ask it to follow that pattern. This is sometimes called "few-shot" prompting.

An example, broken down

Here's a simple prompt:

In one sentence it tells the AI the task (write an email), the length (100 words), the tone (friendly), and the details (the meeting time and what to ask for). That specificity is what makes the output usable.

Why good prompts matter

The same AI can give you a brilliant answer or a useless one depending on the prompt. Because the model takes your wording at face value, small improvements — adding context, naming a format, setting a length — often produce dramatically better results. Learning to prompt well is the single highest-leverage skill for getting value out of AI, and it doesn't require any technical background.

Frequently asked questions