How to beginner · 3 min read

How to set up DSPy with OpenAI

Q: How to set up DSPy with OpenAI

To set up dspy with OpenAI, install the dspy and openai packages, then configure dspy with an OpenAI language model instance using your API key from os.environ. Use dspy.LM to create the model and dspy.configure to set it as the default for declarative AI calls.

Quick answer

To set up dspy with OpenAI, install the dspy and openai packages, then configure dspy with an OpenAI language model instance using your API key from os.environ. Use dspy.LM to create the model and dspy.configure to set it as the default for declarative AI calls.

PREREQUISITES

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

Setup

Install the required packages dspy and openai via pip, and set your OpenAI API key as an environment variable.

bash

pip install dspy openai>=1.0

# Set your OpenAI API key in your shell environment
export OPENAI_API_KEY="your_openai_api_key_here"

Step by step

This example shows how to create a dspy.LM instance with the OpenAI client, configure dspy to use it, define a simple signature, and run a prediction.

python

import os
from openai import OpenAI
import dspy

# Initialize OpenAI client with API key from environment
client = OpenAI(api_key=os.environ["OPENAI_API_KEY"])

# Create a dspy language model wrapping the OpenAI client
lm = dspy.LM("openai/gpt-4o", client=client)

# Configure dspy to use this LM globally
dspy.configure(lm=lm)

# Define a signature for a simple QA task
class QA(dspy.Signature):
    question: str = dspy.InputField()
    answer: str = dspy.OutputField()

# Create a prediction callable
qa = dspy.Predict(QA)

# Run a prediction
result = qa(question="What is DSPy?")
print("Answer:", result.answer)

output

Answer: DSPy is a declarative Python library for AI programming that simplifies calling language models like OpenAI's GPT-4o.

Common variations

Use different OpenAI models by changing the model name in dspy.LM, e.g., "openai/gpt-4o-mini".
For async usage, integrate with async frameworks but dspy primarily supports synchronous calls.
Use other AI providers by creating a compatible client and passing it to dspy.LM.

Troubleshooting

If you get authentication errors, verify your OPENAI_API_KEY environment variable is set correctly.
If dspy calls fail, ensure you have the latest dspy and openai packages installed.
For unexpected output, check your model name and API usage limits.

✅

Key Takeaways

Use dspy.LM with an OpenAI client for clean AI integration.
Always configure dspy with dspy.configure(lm=lm) before making predictions.
Set your OpenAI API key securely via environment variables to avoid hardcoding.
You can switch models easily by changing the model name in dspy.LM.
Check package versions and environment variables if you encounter errors.

Verified 2026-04 · openai/gpt-4o, openai/gpt-4o-mini

Verify ↗