Prompt engineering

Few-Shot Prompting Explained With Examples

Few-shot prompting explained simply: definition, the 4-block structure, 5 copy-paste prompts with examples, and the mistakes that ruin your results.

Autor: Jérémie Chiari

13 września 2026 10 min czytania

Hand typing on a keyboard with code on the screen

You ask ChatGPT a question and get an answer that's technically fine but flat, too long, or badly formatted. You rewrite your instruction three times and the result barely budges. The problem rarely comes from the model: it comes from the fact that you're describing what you want instead of showing it.

Few-shot prompting fixes exactly that. The technique fits in one sentence: before you make your real request, you slip in two to five examples of the result you expect. The AI spots the pattern and reproduces it. It's the fastest way to get a stable format, a consistent tone, and answers you can use without touching them up.

Below you'll find the plain-language definition, the difference between zero-shot, one-shot and few-shot, the four-block structure of a prompt that works, five complete prompts to copy as-is, three ways to practice, and the mistakes that make the method fail. No technical skills required: it all happens in a normal chat window.

What exactly is few-shot prompting?

Few-shot prompting means dropping a few examples of "request / ideal answer" pairs into your prompt, so the AI copies that model instead of improvising its own.

The word shot here means "example." So we talk about prompts with zero examples, one example, or several examples. Nothing more complicated than that.

Technique	Examples provided	When to use it
Zero-shot	0	Simple request, free-form answer: "Summarize this text in 5 lines"
One-shot	1	You want to impose a specific format on a simple task
Few-shot	2 to 5	Tone, style, classification, industry vocabulary, repetitive format

The term spread with the research paper Language Models are Few-Shot Learners, published in 2020, which showed that a large language model could handle a brand-new task purely from a few examples placed in the instruction, with no retraining at all. Today, the technique works on every consumer assistant: ChatGPT, Claude, Gemini, Mistral.

Worth remembering: you are not training the AI. Your examples apply to the current conversation only, a bit like a document template you'd put on a new colleague's desk before handing them a task.

Why a few examples beat a long instruction

An example conveys in three lines what a written instruction would take a whole paragraph to explain, and it leaves no room for interpretation.

Take an instruction like "reply in a professional but warm way, staying concise." Everyone has their own definition of "warm" and "concise." So does the AI. Show it a four-line reply that opens with a thank-you and closes with an open question, and the ambiguity disappears.

In practice, a good example carries six pieces of information at once that you'd never have thought to write down:

the expected length, down to the word;
the exact format (list, table, paragraph, a single sentence);
the tone and the point of view used (you, we, they);
the vocabulary of your industry, including your in-house phrasings;
the internal structure (hook, argument, call to action);
how to handle edge cases (what to do when information is missing).

These models work by continuing patterns: when they see three blocks built on the same logic, producing a fourth one from the same mold becomes the most natural path. It's the direct complement to the fundamentals covered in our complete method for writing a good prompt: context says what, examples show how.

The structure of a few-shot prompt that works

Four blocks, always in the same order: the context, the instruction, the examples labeled identically, then your real request deliberately left hanging.

Block 1, the context. Two lines are enough: who's speaking, to whom, in what setting. "You are the customer service lead at a hiking gear shop."

Block 2, the instruction. What you want to get, in one sentence. "Write a public reply to every customer review posted on Google."

Block 3, the examples. Two to five pairs, all built with the same labels. This is the heart of the technique and it's where most people go wrong: if you write "Review:" in the first example and "Customer comment:" in the second, you break the pattern.

Block 4, the new input. You lay out your real case with the same label, then you write the output label and stop. That blank is a strong signal: the AI understands it has to fill it in.

Review: Fast delivery, but the size doesn't match the guide. Reply: Thanks for the feedback, and glad the delivery hit the mark. We're sorry about the sizing discrepancy: our guide has just been corrected. Write to us at contact@example.com and we'll arrange a free exchange.

Review: Product as described, packaging damaged on arrival. Reply: Thanks for taking the time to write to us. Damaged packaging isn't normal, and we're flagging the issue with the carrier today. If the product itself was affected, reply to this message and we'll replace the item.

Review: Three weeks of waiting and zero response from customer service. Reply:

Two formatting rules make the difference. First, keep your separators rigorously identical from one example to the next. Second, vary the cases: a positive review, a mixed one, a negative one. Anthropic's documentation on multishot prompting recommends three to five relevant and varied examples, precisely so the model grasps the general rule rather than one specific case.

Five few-shot prompting examples to copy

Here are five complete prompts, on tasks you can test in the next ten minutes; just swap in your own content.

1. Replying to customer reviews

That's the example above. Three reviews covering the three possible tones, and you get consistent replies that anyone on the team can publish. Typical gain: from about ten minutes per review down to under a minute of proofreading.

2. Sorting messages by urgency level

Classify each message as URGENT, NORMAL or INFO, then give the action to take in five words maximum.

Message: The order page has been showing a 500 error since 9 a.m. Classification: URGENT | Action: alert the technical team

Message: Can you send me the minutes from Tuesday's meeting? Classification: NORMAL | Action: send the document

Message: The September newsletter went out this morning. Classification: INFO | Action: none

Message: A client is threatening to cancel their contract tomorrow. Classification:

The pipe and the capital letters aren't decorative: they make the output easy to paste into a spreadsheet.

3. Turning a feature into a customer benefit

Feature: 5000 mAh battery Benefit: Two full days without hunting for an outlet

Feature: Recycled aluminum case Benefit: As tough as a standard model, with a smaller footprint

Feature: 120 Hz display Benefit:

Three lines, and the model has understood that you want short, usage-oriented phrasing with no marketing adjectives. A written instruction would have needed a whole paragraph.

4. Extracting information from a text

Extract the information as a table row: Name | Company | Need | Deadline. Write "not specified" if the information is missing.

Text: Hello, I'm Marc Dupuis from Verlaine Ltd, we're looking for AI training for six sales reps before December. Row: Marc Dupuis | Verlaine Ltd | AI training for 6 sales reps | December

Text: Hi, it's Léa, I'd like some information about your pricing. Row: Léa | not specified | pricing information | not specified

Text: [paste your email here] Row:

The second example exists purely to show how to handle missing information. Without it, the AI invents company names. That's the highest-return type of example: the one that covers the awkward case.

5. Writing in your own style

Paste three of your own pieces (three posts, three emails, three intros), each preceded by its topic, then add the topic of the fourth and stop. No description of style will ever beat your own sentences. It's the best way to avoid that smooth, generic writing you can spot from a mile away.

Where to practice few-shot prompting: three methods

The most effective route is to practice on real cases with structured feedback, then experiment freely in the consumer tools.

Method 1: the Skilzy program and AI Lab

Skilzy is a French e-learning platform for learning AI and, above all, actually using it. The prompt fundamentals program covers few-shot step by step, with corrected exercises and reusable templates for your own tasks. What sets it apart is the built-in AI Lab: you write your prompts on the real tools, credits included, without stacking up subscriptions on the side.

The platform has more than fifteen programs (images, video, UGC, faceless channels, avatars, voice-over, music, automation with n8n, AI-assisted coding) and prepares you for two state-recognized certifications eligible for training funding: RS7439 in AI content marketing and RS6792 in AI and sales. Subscriptions start at €29.90 per month with no commitment, and there's a B2B offer for training organizations.

To try before you pay, the discovery demo gives you 7 days with 1 image, 1 video, 1 music track and 10 messages, no credit card (an account is enough).

Method 2: straight into ChatGPT, Claude or Gemini

The free versions of these three assistants are more than enough to practice few-shot. The method: pick a task you repeat every week, write three examples of the perfect output by hand, and compare the result with and without the examples. The gap will convince you in a single session.

Keep the prompts that work in a simple document, sorted by task. A well-built few-shot prompt stays reusable for months: you only change the last block.

Method 3: the vendors' prompt libraries

Anthropic, OpenAI and Google publish freely available prompt libraries and prompt engineering guides. The point isn't to copy the prompts as-is, but to observe how the examples are labeled and separated. One caveat: these resources are mostly in English and written for developers, which makes them less comfortable when you're just starting out.

The mistakes that ruin a few-shot prompt

Most failures come down to three things: examples that don't match each other, examples that are all alike, or mediocre examples that the AI faithfully reproduces.

Changing labels halfway through. "Review:" then "Comment:" is enough to blur the pattern. Copy-paste your structure.
Giving examples that are too similar. Three positive reviews, and the model will reply sweetly to a furious complaint. Cover the extreme cases.
Providing rough examples. The AI imitates what you show it, typos included. One sloppy example produces ten sloppy answers.
Putting in fifteen of them. Past five, you lengthen the prompt without gaining precision, and you risk locking the model onto irrelevant details.
Forgetting the instruction. Examples show the how, not the why. Keep one sentence explaining the goal.
Writing the output in the AI's place. End on the empty label. If you start the answer, you'll get a variation of your own sentence.

These habits overlap with the traps listed in our article on prompting mistakes that wreck your results. And when the task calls for reasoning rather than formatting (calculation, analysis, decision), combine few-shot with chain-of-thought, which makes the AI reason step by step: you then show not just the answer, but the path to reach it.

Key takeaways

Showing costs less than explaining. Two to five well-chosen examples, labeled the same way, varied, and ending on a blank to fill in: that's the whole technique. Take a task you repeat every week, build the prompt once, reuse it a hundred times.

Najczęściej zadawane pytania

How many examples should you give in a few-shot prompt?

Between two and five in the vast majority of cases. A single example is enough to impose a simple format. Three is the right default setting: one typical case, one difficult case, and one case where information is missing. Past five, you're lengthening the prompt without gaining quality, and you even risk making the AI too rigid, focused on details from your examples rather than on the general rule.

What's the difference between zero-shot and few-shot prompting?

In zero-shot, you give an instruction with no examples at all: "Summarize this text in five points." In few-shot, you add several examples of the expected result before your real request. Zero-shot is faster and works well for open-ended tasks. Few-shot becomes clearly superior as soon as a precise format, a particular tone, or industry vocabulary comes into play, and as soon as you have to repeat the same task several times with a consistent result.

Does few-shot prompting work with the free version of ChatGPT?

Yes. The technique doesn't depend on any paid feature: it rests entirely on how you write your message. It works with the free versions of ChatGPT, Claude, Gemini or Mistral, in the usual chat window. Paid versions mainly bring longer context windows and more capable models, which helps when your examples are lengthy, but it's not a prerequisite for getting started.

Do my examples actually train the AI?

No, and this is a common confusion. Your examples don't train the model: they steer it for the duration of the current conversation. If you open a new chat, you have to provide them again. That's why it's worth keeping the few-shot prompts that work in a document, sorted by task. Some tools offer custom assistants where you save those examples once and for all.

Can you combine few-shot with other prompting techniques?

Yes, and it's actually recommended. The most useful combination pairs few-shot with chain-of-thought: in each example, you show not just the final answer but the reasoning that leads to it. The model then reproduces the approach, which clearly improves analysis, calculation and decision tasks. You can also add a role at the start of the prompt and format constraints at the end, without ever touching the structure of your examples.

How do I know if my few-shot prompt is working?

Run a two-step test: send your request without examples, then with them. If the second answer is directly usable with no tweaks to format or tone, you've nailed it. If the result stays uneven from one attempt to the next, the problem almost always comes from the examples: they're too similar to each other, they don't use the same labels, or there's no case showing what to do when a piece of information is missing.