How to Beginner to Intermediate · 3 min read

How to prompt AI to debug code

Quick answer

To prompt AI to debug code, provide the code snippet along with a clear instruction like "Find and fix bugs in this code." Use explicit context such as the programming language and expected behavior. For example, prompt with: "Debug this Python function that calculates factorial."

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

Use the gpt-4o model with a prompt that includes the buggy code and a clear debugging instruction. The example below shows a complete script that sends a prompt to the OpenAI API and prints the debugged code.

python

import os
from openai import OpenAI

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

buggy_code = '''
def factorial(n):
    if n == 0:
        return 0
    else:
        return n * factorial(n-1)
'''

prompt = f"Debug this Python function and fix any errors:\n{buggy_code}"

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

print("Debugged code:\n", response.choices[0].message.content)

output

Debugged code:
 def factorial(n):
    if n == 0:
        return 1
    else:
        return n * factorial(n-1)

Common variations

You can use other models like claude-3-5-sonnet-20241022 or enable streaming for real-time output. Async calls are also supported in the latest SDKs.

python

import os
from anthropic import Anthropic

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

buggy_code = '''
def add_numbers(a, b):
    return a - b
'''

system_prompt = "You are a helpful assistant that debugs code."
user_prompt = f"Debug this Python function and fix any errors:\n{buggy_code}"

message = client.messages.create(
    model="claude-3-5-sonnet-20241022",
    max_tokens=200,
    system=system_prompt,
    messages=[{"role": "user", "content": user_prompt}]
)

print("Debugged code:\n", message.content[0].text)

output

Debugged code:
 def add_numbers(a, b):
    return a + b

Troubleshooting

If the AI returns unrelated text, clarify the prompt by adding "Only provide corrected code without explanations."
If the output is incomplete, increase max_tokens or use streaming.
For ambiguous bugs, provide expected input/output examples in the prompt.

✅

Key Takeaways

Always include the buggy code and a clear debugging instruction in your prompt.
Specify the programming language and expected behavior to guide the AI.
Use the latest SDK patterns with environment variables for API keys.
Try different models like gpt-4o or claude-3-5-sonnet-20241022 for best debugging results.
Add constraints in the prompt to avoid verbose or unrelated outputs.

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

Verify ↗