Langfuse

Langfuse is an open-source LLM observability platform that provides prompt management, analytics, and evaluation.

Features

Prompt tracing - Log all prompts and responses.
Cost tracking - Monitor token usage and costs.
Latency analytics - Track response times.
Prompt management - Version and deploy prompts.
Evaluation - Score and evaluate outputs.
User tracking - Attribute usage to users.

Setup

Self-hosted Langfuse

Deploy Langfuse to your Kubernetes cluster.

# Create namespace
kubectl create namespace langfuse

# Deploy PostgreSQL
kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
  name: langfuse-db
  namespace: langfuse
spec:
  ports:
  - port: 5432
  selector:
    app: langfuse-db
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: langfuse-db
  namespace: langfuse
spec:
  selector:
    matchLabels:
      app: langfuse-db
  template:
    metadata:
      labels:
        app: langfuse-db
    spec:
      containers:
      - name: postgres
        image: postgres:15
        env:
        - name: POSTGRES_USER
          value: langfuse
        - name: POSTGRES_PASSWORD
          value: langfuse
        - name: POSTGRES_DB
          value: langfuse
        ports:
        - containerPort: 5432
EOF

# Deploy Langfuse
kubectl apply -f - <<EOF
apiVersion: v1
kind: Service
metadata:
  name: langfuse
  namespace: langfuse
spec:
  ports:
  - port: 3000
    targetPort: 3000
  selector:
    app: langfuse
---
apiVersion: apps/v1
kind: Deployment
metadata:
  name: langfuse
  namespace: langfuse
spec:
  selector:
    matchLabels:
      app: langfuse
  template:
    metadata:
      labels:
        app: langfuse
    spec:
      containers:
      - name: langfuse
        image: langfuse/langfuse:latest
        env:
        - name: DATABASE_URL
          value: postgresql://langfuse:langfuse@langfuse-db:5432/langfuse
        - name: NEXTAUTH_SECRET
          value: your-secret-change-me
        - name: NEXTAUTH_URL
          value: http://langfuse.langfuse.svc.cluster.local:3000
        ports:
        - containerPort: 3000
EOF

Access Langfuse by port-forwarding.

kubectl port-forward -n langfuse svc/langfuse 3000:3000

Then navigate to http://localhost:3000.

Cloud Langfuse

Configuration

Configure the OpenTelemetry Collector to forward traces to Langfuse.

# Update the traces collector
helm upgrade --install opentelemetry-collector-traces opentelemetry-collector \
  --repo https://open-telemetry.github.io/opentelemetry-helm-charts \
  --version 0.127.2 \
  --set mode=deployment \
  --set image.repository="otel/opentelemetry-collector-contrib" \
  --set command.name="otelcol-contrib" \
  --namespace=telemetry \
  --create-namespace \
  -f -<<EOF
config:
  receivers:
    otlp:
      protocols:
        grpc:
          endpoint: 0.0.0.0:4317
        http:
          endpoint: 0.0.0.0:4318
  exporters:
    # Export to Langfuse Cloud
    otlphttp/langfuse:
      endpoint: https://cloud.langfuse.com/api/public/otel
      headers:
        Authorization: "Basic <base64-encoded-public-key:secret-key>"
    # Or export to self-hosted Langfuse
    otlphttp/langfuse-local:
      endpoint: http://langfuse.langfuse.svc.cluster.local:3000/api/public/otel
    debug:
      verbosity: detailed
  service:
    pipelines:
      traces:
        receivers: [otlp]
        exporters: [debug, otlphttp/langfuse]  # Use otlphttp/langfuse-local for self-hosted
EOF

ℹ️

To create the base64-encoded credentials for Langfuse Cloud, run the following command.

echo -n "public-key:secret-key" | base64

Verify integration

Send a request through agentgateway to an LLM backend.

curl -X POST http://localhost:8080/v1/chat/completions \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-4o-mini",
    "messages": [{"role": "user", "content": "Hello!"}]
  }'

Navigate to your Langfuse dashboard and verify that the trace appears with the following information.
- Full prompt and response.
- Token counts (input and output).
- Model information.
- Latency metrics.

Learn more

LangSmith