How to Beginner · 3 min read

How to build AI email categorizer

Q: How to build AI email categorizer

Use a large language model like gpt-4o to classify emails by sending the email text as input and prompting the model to assign categories. Implement this by calling client.chat.completions.create with a prompt that instructs the model to categorize the email content.

Quick answer

Use a large language model like gpt-4o to classify emails by sending the email text as input and prompting the model to assign categories. Implement this by calling client.chat.completions.create with a prompt that instructs the model to categorize the email content.

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 for secure access.

bash

pip install openai>=1.0

Step by step

This example shows how to send an email text to gpt-4o and get a category label such as "Spam", "Promotion", or "Personal".

python

import os
from openai import OpenAI

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

email_text = """Subject: Huge discounts on electronics!

Dear customer, check out our exclusive deals on laptops and phones. Limited time offer!"""

prompt = f"Categorize the following email into one of these categories: Spam, Promotion, Personal, Work.\n\nEmail:\n{email_text}\n\nCategory:" 

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

category = response.choices[0].message.content.strip()
print(f"Email category: {category}")

output

Email category: Promotion

Common variations

Use claude-3-5-sonnet-20241022 for potentially better classification accuracy.
Implement async calls with asyncio for batch processing.
Use streaming completions for real-time categorization feedback.
Expand categories or use embeddings + vector search for more complex classification.

Troubleshooting

If you get empty or irrelevant categories, refine the prompt to be more explicit.
Check your API key environment variable if authentication errors occur.
For rate limits, implement exponential backoff retries.
Ensure email text is clean and well formatted for best results.

✅

Key Takeaways

Use gpt-4o with clear prompts to categorize emails effectively.
Always keep your API key secure using environment variables.
Experiment with different models like claude-3-5-sonnet-20241022 for improved accuracy.
Prompt engineering is critical to get precise category labels.
Handle API errors and rate limits gracefully in production.

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

Verify ↗