How to beginner · 3 min read

Langfuse database setup

Quick answer

To set up a Langfuse database, install the langfuse Python package, configure your database connection string in the Langfuse client, and initialize it with your API keys. Use Langfuse to automatically trace and store AI call data in your database for observability and analysis.

PREREQUISITES

Python 3.8+
Langfuse account with API keys
pip install langfuse
Access to a supported database (PostgreSQL, MySQL, SQLite, etc.)

Setup

Install the langfuse Python package and set your environment variables for API keys. Prepare your database connection string (DSN) for Langfuse to store trace data.

bash

pip install langfuse

Step by step

Initialize the Langfuse client with your public and secret keys and configure the database connection string. Use the client to trace AI calls, which will be saved in your database.

python

import os
from langfuse import Langfuse

# Set environment variables before running this script:
# export LANGFUSE_PUBLIC_KEY=os.environ["LANGFUSE_PUBLIC_KEY"]
# export LANGFUSE_SECRET_KEY=os.environ["LANGFUSE_SECRET_KEY"]

langfuse = Langfuse(
    public_key=os.environ["LANGFUSE_PUBLIC_KEY"],
    secret_key=os.environ["LANGFUSE_SECRET_KEY"],
    host="https://cloud.langfuse.com",
    database_url="postgresql://user:password@localhost:5432/langfuse_db"  # Example DSN
)

# Example function to trace an AI call
@langfuse.observe()
def call_ai_model(prompt: str) -> str:
    # Simulate AI call
    response = f"Response to: {prompt}"
    return response

result = call_ai_model("Hello Langfuse")
print("AI response:", result)

output

AI response: Response to: Hello Langfuse

Common variations

Use async functions with @langfuse.observe() for asynchronous AI calls.
Configure different database backends by changing the database_url parameter (e.g., MySQL, SQLite).
Integrate Langfuse with OpenAI or other AI SDKs by wrapping their calls with @observe() decorators.

python

import asyncio

@langfuse.observe()
async def async_call_ai(prompt: str) -> str:
    await asyncio.sleep(0.1)  # Simulate async call
    return f"Async response to: {prompt}"

async def main():
    response = await async_call_ai("Async Langfuse")
    print(response)

asyncio.run(main())

output

Async response to: Async Langfuse

Troubleshooting

If you see connection errors, verify your database_url and network access to the database.
Ensure your LANGFUSE_PUBLIC_KEY and LANGFUSE_SECRET_KEY environment variables are set correctly.
Check Langfuse logs for detailed error messages and enable debug logging if needed.

✅

Key Takeaways

Install and configure the langfuse Python package with your API keys and database URL.
Use the @observe() decorator to automatically trace AI calls and store data in your database.
Support for multiple databases allows flexible backend choices for Langfuse observability.
Async and sync AI calls can both be traced seamlessly with Langfuse.
Verify environment variables and database connectivity to avoid common setup issues.

Verified 2026-04

Verify ↗