How to beginner · 4 min read

Batch classification with OpenAI API

Q: Batch classification with OpenAI API

Use the OpenAI API's chat.completions.create method with a batch of classification prompts by sending multiple messages in a single request or iterating over inputs. Batch processing improves throughput and reduces latency when classifying many texts.

Quick answer

Use the OpenAI API's chat.completions.create method with a batch of classification prompts by sending multiple messages in a single request or iterating over inputs. Batch processing improves throughput and reduces latency when classifying many texts.

PREREQUISITES

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

Setup

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

bash

pip install openai>=1.0

output

Collecting openai
  Downloading openai-1.x.x-py3-none-any.whl (xx kB)
Installing collected packages: openai
Successfully installed openai-1.x.x

Step by step

This example demonstrates batch classification by sending multiple user messages in one chat.completions.create call using the gpt-4o model. Each message contains a text to classify with a prompt instructing the model to output the category.

python

import os
from openai import OpenAI

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

texts_to_classify = [
    "I love this product!",
    "The movie was boring and too long.",
    "Great customer service and fast shipping.",
    "The food was cold and tasteless."
]

messages = [
    {
        "role": "user",
        "content": f"Classify the sentiment of this text as Positive, Negative, or Neutral:\n\n{txt}"
    }
    for txt in texts_to_classify
]

response = client.chat.completions.create(
    model="gpt-4o",
    messages=messages
)

for i, choice in enumerate(response.choices):
    print(f"Text: {texts_to_classify[i]}")
    print(f"Classification: {choice.message.content.strip()}")
    print("---")

output

Text: I love this product!
Classification: Positive
---
Text: The movie was boring and too long.
Classification: Negative
---
Text: Great customer service and fast shipping.
Classification: Positive
---
Text: The food was cold and tasteless.
Classification: Negative
---

Common variations

Async calls: Use async and await with the OpenAI client for concurrency.
Streaming: Stream partial classification results with stream=True.
Different models: Use gpt-4o-mini for faster, cheaper classification or claude-3-5-sonnet-20241022 for Anthropic API.

python

import asyncio
import os
from openai import OpenAI

async def batch_classify_async(texts):
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    messages = [
        {"role": "user", "content": f"Classify the sentiment of this text as Positive, Negative, or Neutral:\n\n{txt}"}
        for txt in texts
    ]
    response = await client.chat.completions.acreate(
        model="gpt-4o-mini",
        messages=messages
    )
    for i, choice in enumerate(response.choices):
        print(f"Text: {texts[i]}")
        print(f"Classification: {choice.message.content.strip()}")
        print("---")

if __name__ == "__main__":
    texts = [
        "I love this product!",
        "The movie was boring and too long.",
        "Great customer service and fast shipping.",
        "The food was cold and tasteless."
    ]
    asyncio.run(batch_classify_async(texts))

output

Text: I love this product!
Classification: Positive
---
Text: The movie was boring and too long.
Classification: Negative
---
Text: Great customer service and fast shipping.
Classification: Positive
---
Text: The food was cold and tasteless.
Classification: Negative
---

Troubleshooting

If you receive RateLimitError, reduce batch size or add retry logic.
If classifications are inconsistent, refine your prompt for clarity.
Ensure your API key is set correctly in OPENAI_API_KEY environment variable.

✅

Key Takeaways

Batch classification improves throughput by sending multiple inputs in one API call.
Use clear, consistent prompts to get reliable classification results.
Async calls enable concurrent batch processing for better performance.
Adjust model choice based on speed, cost, and accuracy needs.
Handle rate limits by controlling batch size and implementing retries.

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

Verify ↗