How to beginner · 3 min read

How to use few-shot examples in prompts

Q: How to use few-shot examples in prompts

Use few-shot examples by including a few input-output pairs in your prompt before the new query to guide the model's behavior. This technique helps models like gpt-4o understand the task format and produce more accurate responses.

Quick answer

Use few-shot examples by including a few input-output pairs in your prompt before the new query to guide the model's behavior. This technique helps models like gpt-4o understand the task format and produce more accurate responses.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install openai>=1.0

Setup

Install the OpenAI Python SDK and set your API key as an environment variable.

bash

pip install openai>=1.0

Step by step

Include a few input-output examples in the prompt to demonstrate the task, then add the new input for the model to complete.

python

import os
from openai import OpenAI

client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

few_shot_prompt = '''
Translate English to French:

English: Hello, how are you?
French: Bonjour, comment ça va ?

English: What is your name?
French: Comment vous appelez-vous ?

English: Where is the library?
French:'''

response = client.chat.completions.create(
    model="gpt-4o",
    messages=[{"role": "user", "content": few_shot_prompt}]
)

print(response.choices[0].message.content.strip())

output

Où est la bibliothèque ?

Common variations

You can use few-shot prompting with different models like claude-3-5-haiku-20241022 or asynchronously. Adjust examples to fit your task and model capabilities.

python

import anthropic
import os

client = anthropic.Anthropic(api_key=os.environ["ANTHROPIC_API_KEY"])

few_shot_prompt = '''Translate English to French:

English: Hello, how are you?
French: Bonjour, comment ça va ?

English: What is your name?
French: Comment vous appelez-vous ?

English: Where is the library?
French:'''  

message = client.messages.create(
    model="claude-3-5-haiku-20241022",
    max_tokens=100,
    system="You are a helpful assistant.",
    messages=[{"role": "user", "content": few_shot_prompt}]
)

print(message.content[0].text.strip())

output

Où est la bibliothèque ?

Troubleshooting

If the model output is off-topic or incomplete, ensure your few-shot examples are clear, consistent, and closely related to the task. Also, keep the prompt concise to avoid token limits.

✅

Key Takeaways

Include clear input-output pairs before your query to guide the model with few-shot examples.
Use the latest SDK patterns and models like gpt-4o or claude-3-5-haiku-20241022 for best results.
Keep examples relevant and concise to avoid exceeding token limits and confusing the model.

Verified 2026-04 · gpt-4o, claude-3-5-haiku-20241022

Verify ↗