Langfuse Install Auth

Langfuse Install & Auth

Overview

Set up Langfuse SDK and configure authentication for LLM observability and tracing.

Prerequisites

• Node.js 18+ or Python 3.9+
• Package manager (npm, pnpm, or pip)
• Langfuse account (cloud or self-hosted)
• Public and Secret keys from Langfuse dashboard

Instructions

Step 1: Install SDK

set -euo pipefail
# Node.js / TypeScript
npm install langfuse
# or
pnpm add langfuse

# Python
pip install langfuse

Step 2: Get API Keys

1. Go to Langfuse dashboard (https://cloud.langfuse.com or your self-hosted instance)
2. Navigate to Settings -> API Keys
3. Create new API key pair (Public Key + Secret Key)
4. Note your host URL (cloud: https://cloud.langfuse.com)

Step 3: Configure Authentication

# Set environment variables
export LANGFUSE_PUBLIC_KEY="pk-lf-..."
export LANGFUSE_SECRET_KEY="sk-lf-..."
export LANGFUSE_HOST="https://cloud.langfuse.com"  # or self-hosted URL

# Or create .env file
cat >> .env << 'EOF'
LANGFUSE_PUBLIC_KEY=pk-lf-your-public-key
LANGFUSE_SECRET_KEY=sk-lf-your-secret-key
LANGFUSE_HOST=https://cloud.langfuse.com
EOF

Step 4: Verify Connection

// TypeScript verification
import { Langfuse } from "langfuse";

const langfuse = new Langfuse({
  publicKey: process.env.LANGFUSE_PUBLIC_KEY,
  secretKey: process.env.LANGFUSE_SECRET_KEY,
  baseUrl: process.env.LANGFUSE_HOST,
});

// Create a test trace
const trace = langfuse.trace({
  name: "connection-test",
  metadata: { test: true },
});

// Flush to ensure data is sent
await langfuse.flushAsync();
console.log("Langfuse connection successful! Check dashboard for trace.");

# Python verification
from langfuse import Langfuse
import os

langfuse = Langfuse(
    public_key=os.environ.get("LANGFUSE_PUBLIC_KEY"),
    secret_key=os.environ.get("LANGFUSE_SECRET_KEY"),
    host=os.environ.get("LANGFUSE_HOST"),
)

# Create a test trace
trace = langfuse.trace(name="connection-test", metadata={"test": True})

# Flush to ensure data is sent
langfuse.flush()
print("Langfuse connection successful! Check dashboard for trace.")

Output

• Installed SDK package in node_modules or site-packages
• Environment variables or .env file with API credentials
• Successful connection verification trace visible in dashboard

Error Handling

Examples

TypeScript Initialization with Options

import { Langfuse } from "langfuse";

const langfuse = new Langfuse({
  publicKey: process.env.LANGFUSE_PUBLIC_KEY!,
  secretKey: process.env.LANGFUSE_SECRET_KEY!,
  baseUrl: process.env.LANGFUSE_HOST,
  // Optional configuration
  flushAt: 15,        // Flush after 15 events (default: 15)
  flushInterval: 10000, // Flush every 10 seconds (default: 10000ms)  # 10000: 10 seconds in ms
  requestTimeout: 10000, // Request timeout (default: 10000ms)  # 10 seconds in ms
  enabled: process.env.NODE_ENV === "production", // Disable in dev
});

// Ensure clean shutdown
process.on("beforeExit", async () => {
  await langfuse.shutdownAsync();
});

Python Initialization with Options

from langfuse import Langfuse
import os

langfuse = Langfuse(
    public_key=os.environ.get("LANGFUSE_PUBLIC_KEY"),
    secret_key=os.environ.get("LANGFUSE_SECRET_KEY"),
    host=os.environ.get("LANGFUSE_HOST"),
    # Optional configuration
    flush_at=15,           # Flush after 15 events
    flush_interval=10.0,   # Flush every 10 seconds
    enabled=os.environ.get("NODE_ENV") == "production",
)

# Register shutdown handler
import atexit
atexit.register(langfuse.flush)

OpenAI Integration (Automatic Tracing)

import { observeOpenAI } from "langfuse";
import OpenAI from "openai";

const openai = observeOpenAI(new OpenAI(), {
  clientInitParams: {
    publicKey: process.env.LANGFUSE_PUBLIC_KEY,
    secretKey: process.env.LANGFUSE_SECRET_KEY,
    baseUrl: process.env.LANGFUSE_HOST,
  },
});

// All OpenAI calls are now automatically traced
const response = await openai.chat.completions.create({
  model: "gpt-4",
  messages: [{ role: "user", content: "Hello!" }],
});

Resources

• Langfuse Documentation
• Langfuse SDK Reference
• Langfuse Cloud Dashboard
• Self-Hosting Guide

Next Steps

After successful auth, proceed to langfuse-hello-world for your first trace.