How to beginner · 3 min read

Prompt engineering for code generation

Quick answer

Use clear, specific prompts with context and examples to guide LLMs like gpt-4o for code generation. Include instructions on language, style, and output format to get precise, runnable code.

PREREQUISITES

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

Setup

Install the openai Python package and set your API key as an environment variable for secure access.

bash

pip install openai>=1.0

Step by step

This example shows how to craft a prompt for generating Python code that calculates Fibonacci numbers. The prompt specifies language, function name, and output format.

python

import os
from openai import OpenAI

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

prompt = (
    "Write a Python function named fibonacci that returns the nth Fibonacci number. "
    "Include type hints and a docstring explaining the function. "
    "Return only the code without extra explanation."
)

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

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

output

def fibonacci(n: int) -> int:
    """Return the nth Fibonacci number."""
    if n <= 0:
        return 0
    elif n == 1:
        return 1
    else:
        a, b = 0, 1
        for _ in range(2, n + 1):
            a, b = b, a + b
        return b

Common variations

You can use async calls, streaming responses, or switch models like gpt-4o-mini for faster, cheaper code generation. Adjust prompts to specify language or coding style.

python

import asyncio
import os
from openai import OpenAI

async def async_code_generation():
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    prompt = (
        "Generate a JavaScript function to reverse a string. "
        "Return only the code."
    )
    stream = await client.chat.completions.create(
        model="gpt-4o-mini",
        messages=[{"role": "user", "content": prompt}],
        stream=True
    )
    async for chunk in stream:
        print(chunk.choices[0].delta.content or "", end="", flush=True)

if __name__ == "__main__":
    asyncio.run(async_code_generation())

output

function reverseString(str) {
    return str.split('').reverse().join('');
}

Troubleshooting

If the generated code is incomplete, increase max_tokens in your request.
If the output includes explanations, clarify in the prompt to return only code.
For syntax errors, specify the programming language explicitly in the prompt.

✅

Key Takeaways

Be explicit and detailed in your prompt to get accurate code output.
Use examples or specify output format to reduce ambiguity.
Leverage streaming and async calls for efficient code generation workflows.

Verified 2026-04 · gpt-4o-mini

Verify ↗