---
title: "Deploy Langfuse v4.16.0 (Production ready AI Observability)"
description: "Open-source LLM platform for prompt management, tracing, evals & metrics"
category: "AI/ML"
url: https://railway.com/deploy/langfuse-v410-production-ready-ai-observ
---

# Deploy Langfuse v4.16.0 (Production ready AI Observability)

Open-source LLM platform for prompt management, tracing, evals & metrics

**[Deploy Langfuse v4.16.0 (Production ready AI Observability) on Railway](https://railway.com/template/langfuse-v410-production-ready-ai-observ)**

- **Creator:** RailGenius
- **Category:** AI/ML
- **Total deploys:** 4

## Template content

### redis https://devicons.railway.app/i/redis.svg

- **Image:** redis:8.2.1

### langfuse-worker https://langfuse.com/icon.svg

- **Image:** langfuse/langfuse-worker:4.16.0

### postgres https://devicons.railway.app/i/postgresql.svg

- **Image:** ghcr.io/railwayapp-templates/postgres-ssl:16.11

### clickhouse https://devicons.railway.app/ClickHouse?variant=light

- **Image:** clickhouse/clickhouse-server:26.4
- **Health check:** /ping

### minio https://res.cloudinary.com/h5kvzw/image/upload/v1706062979/MINIO_Bird_vbtpet.png

- **Image:** minio/minio
- **Start command:** `sh -c 'mkdir -p /data/langfuse && minio server --address ":9000" --console-address ":9001" /data'`
- **Health check:** /minio/health/ready

### langfuse-web https://langfuse.com/icon.svg

- **Image:** langfuse/langfuse:4.16.0
- **Public domain:** Yes

## Documentation

# Deploy and Host Langfuse v4 with Railway

Langfuse is an open-source LLM engineering platform that helps teams collaboratively debug, analyze, and iterate on their LLM applications. This template deploys a production-ready Langfuse v4 stack on Railway with web, worker, and the required data services.

## About Hosting Langfuse

Hosting Langfuse v4 means running two application containers—`langfuse-web` and `langfuse-worker`—plus four infrastructure dependencies: PostgreSQL for transactional data, ClickHouse for high-volume observability analytics, Redis for caching and queues, and S3-compatible object storage (MinIO) for event uploads and large payloads. Traces are ingested via the web service, queued through Redis/object storage, then processed by the worker into ClickHouse. Railway wires private networking and volumes so the full stack comes up as one connected project with a public URL for the Langfuse UI and API.

## Why Deploy Langfuse on Railway

Railway is a singular platform to deploy your infrastructure stack. Railway will host your infrastructure so you don't have to deal with configuration, while allowing you to vertically and horizontally scale it.

By deploying Langfuse on Railway, you are one step closer to supporting a complete full-stack LLM application with minimal burden. Host your observability backend, databases, workers, object storage, and related AI tooling on one platform instead of stitching multiple cloud services together.

## Common Use Cases

- **LLM application observability** — Instrument your app and ingest traces to track LLM calls plus retrieval, embedding, and agent actions; inspect complex logs and user sessions.
- **Prompt management** — Centrally manage, version, and collaboratively iterate on prompts with strong server/client caching so updates don’t add latency.
- **Evaluations** — Run LLM-as-a-judge, collect user feedback, label manually, or build custom eval pipelines via APIs/SDKs.
- **Datasets & experiments** — Maintain test sets and benchmarks for pre-deployment testing, continuous improvement, and framework integrations (LangChain, LlamaIndex, and more).
- **LLM playground & custom LLMOps** — Iterate on prompts and model configs from bad traces, or power bespoke workflows using Langfuse’s API, OpenAPI spec, and typed Python / JS/TS SDKs.

## Dependencies for Langfuse Hosting

This template includes the full Langfuse v4 self-hosted architecture:

- **Langfuse Web** (`langfuse/langfuse:4`) — UI, API, and ingestion endpoint
- **Langfuse Worker** (`langfuse/langfuse-worker:4`) — async processing of queued traces and jobs
- **PostgreSQL** (≥ 15, 16 recommended) — OLTP / transactional data (users, projects, prompts, datasets)
- **ClickHouse** (≥ 25.12, 26.4 recommended for v4) — OLAP store for traces, observations, and scores
- **Redis** (≥ 7.0, 7.2 recommended) — cache and queue (prefer official `redis:7` over Bitnami on Railway)
- **MinIO** (S3-compatible) — blob storage for event uploads, multi-modal inputs, and large exports

### Deployment Dependencies

- [Langfuse self-hosting overview](https://langfuse.com/self-hosting)
- [Langfuse v4 containers (web & worker)](https://langfuse.com/self-hosting/deployment/infrastructure/containers)
- [Upgrade guide: Langfuse v3 → v4](https://langfuse.com/self-hosting/upgrade/upgrade-guides/upgrade-v3-to-v4)
- [Langfuse on Railway (docs)](https://langfuse.com/self-hosting/deployment/railway)
- [Railway: Langfuse LLM observability guide](https://docs.railway.com/guides/langfuse-llm-observability)
- [ClickHouse requirements for v4](https://langfuse.com/self-hosting/deployment/infrastructure/clickhouse)
- [Langfuse GitHub](https://github.com/langfuse/langfuse)
- [Langfuse documentation](https://langfuse.com/docs)

### Implementation Details

After deploy:

Open the public URL on **langfuse-web** and create your first account/org.

## Similar templates

- [Chat Chat](https://railway.com/deploy/-WWW5r) — Chat Chat, your own unified chat and search to AI platform.
- [stella](https://railway.com/deploy/stella) — Self-host stella with web, API, Postgres, Redis, and object storage.
- [Hermes Agent | OpenClaw Alternative with Dashboard](https://railway.com/deploy/hermes-agent-or-openclaw-alternative-wit) — Self-Hosted Hermes AI Agent for Telegram, Discord & Slack

Open this page in a browser: https://railway.com/deploy/langfuse-v410-production-ready-ai-observ