How to beginner · 3 min read

How to create dynamic prompts in LangChain

Q: How to create dynamic prompts in LangChain

Use ChatPromptTemplate from langchain_core.prompts to define prompt templates with variables, then call format to dynamically inject values. Combine with ChatOpenAI to send these dynamic prompts to models like gpt-4o.

Quick answer

Use ChatPromptTemplate from langchain_core.prompts to define prompt templates with variables, then call format to dynamically inject values. Combine with ChatOpenAI to send these dynamic prompts to models like gpt-4o.

PREREQUISITES

Python 3.8+
OpenAI API key (free tier works)
pip install langchain_openai>=0.2 openai>=1.0

Setup

Install the required packages and set your OpenAI API key in the environment.

Install LangChain OpenAI integration and OpenAI SDK:

bash

pip install langchain_openai openai

Step by step

Define a prompt template with variables using ChatPromptTemplate, then format it dynamically and send it to the gpt-4o model using ChatOpenAI.

python

import os
from langchain.prompts import ChatPromptTemplate
from langchain_openai import ChatOpenAI

# Define a prompt template with placeholders
prompt_template = ChatPromptTemplate.from_template(
    "Write a short story about a {adjective} {animal} who loves {activity}."
)

# Format the prompt dynamically
formatted_prompt = prompt_template.format(adjective="brave", animal="fox", activity="dancing")

# Initialize the OpenAI chat client
chat = ChatOpenAI(model_name="gpt-4o", openai_api_key=os.environ["OPENAI_API_KEY"])

# Send the dynamic prompt to the model
response = chat(formatted_prompt)

print("Prompt sent:", formatted_prompt)
print("Model response:", response.content)

output

Prompt sent: Write a short story about a brave fox who loves dancing.
Model response: Once upon a time, a brave fox named Felix loved to dance under the moonlight...

Common variations

You can create dynamic prompts with multiple variables, use async calls, or switch models easily.

Use format_prompt for more complex prompt objects.
Switch to gpt-4o-mini for faster, cheaper calls.
Use async methods with await chat.acall() in async functions.

python

import asyncio

async def async_example():
    prompt_template = ChatPromptTemplate.from_template(
        "Describe a {color} sunset over the {place}."
    )
    prompt = prompt_template.format(color="golden", place="mountains")

    chat = ChatOpenAI(model_name="gpt-4o-mini", openai_api_key=os.environ["OPENAI_API_KEY"])
    response = await chat.acall(prompt)
    print("Async response:", response.content)

asyncio.run(async_example())

output

Async response: The golden sunset painted the mountains with hues of orange and pink, creating a breathtaking view.

Troubleshooting

If you get errors like KeyError when formatting prompts, ensure all variables in the template are provided in format. If API calls fail, verify your OPENAI_API_KEY is set correctly in the environment.

✅

Key Takeaways

Use ChatPromptTemplate to create reusable prompt templates with variables.
Call format on the template to inject dynamic values before sending to the model.
Combine with ChatOpenAI to send dynamic prompts to gpt-4o or other models.
Async calls and different models like gpt-4o-mini are supported for flexibility.
Always provide all required variables to avoid formatting errors and set your API key in os.environ.

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

Verify ↗