How to beginner · 3 min read

How to classify sentiment with LLM

Q: How to classify sentiment with LLM

Use a large language model like gpt-4o via the OpenAI Python SDK to classify sentiment by prompting it with a sentiment analysis task. Send the text as a user message and parse the model's response for sentiment labels such as positive, negative, or neutral.

Quick answer

Use a large language model like gpt-4o via the OpenAI Python SDK to classify sentiment by prompting it with a sentiment analysis task. Send the text as a user message and parse the model's response for sentiment labels such as positive, negative, or neutral.

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.

Install package: pip install openai
Set environment variable in your shell: export OPENAI_API_KEY='your_api_key' (Linux/macOS) or setx OPENAI_API_KEY "your_api_key" (Windows)

bash

pip install openai

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 uses the gpt-4o model to classify sentiment by sending a prompt that instructs the model to label the sentiment of the input text. The response is parsed to extract the sentiment label.

python

import os
from openai import OpenAI

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

text_to_classify = "I love the new design of your website!"

messages = [
    {"role": "system", "content": "You are a sentiment analysis assistant. Respond with one word: Positive, Negative, or Neutral."},
    {"role": "user", "content": f"Classify the sentiment of this text: '{text_to_classify}'"}
]

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

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

output

Sentiment: Positive

Common variations

You can classify sentiment asynchronously, use streaming for real-time output, or switch to other models like gpt-4o-mini or claude-3-5-sonnet-20241022 from Anthropic. For Anthropic, use the system= parameter instead of a system message.

python

import os
import asyncio
from openai import OpenAI

async def classify_sentiment_async(text: str) -> str:
    client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])
    messages = [
        {"role": "system", "content": "You are a sentiment analysis assistant. Respond with Positive, Negative, or Neutral."},
        {"role": "user", "content": f"Classify the sentiment of this text: '{text}'"}
    ]
    response = await client.chat.completions.create(
        model="gpt-4o-mini",
        messages=messages
    )
    return response.choices[0].message.content.strip()

async def main():
    sentiment = await classify_sentiment_async("The movie was okay, not great but not bad.")
    print(f"Sentiment: {sentiment}")

asyncio.run(main())

output

Sentiment: Neutral

Troubleshooting

If you get an authentication error, verify your OPENAI_API_KEY environment variable is set correctly.
If the model returns unexpected output, ensure your system prompt clearly instructs the model to respond with a single sentiment label.
For rate limit errors, consider retrying after a delay or upgrading your API plan.

✅

Key Takeaways

Use the OpenAI SDK with gpt-4o for accurate sentiment classification.
Provide clear instructions in the system prompt to get consistent sentiment labels.
Async and streaming calls enable flexible integration in real-time applications.

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

Verify ↗