Self-host Langfuse on a VPS
Trace, evaluate, and debug LLM applications with prompt management and cost analytics.
levelrail-cli templates deploy langfuse --name my-langfuseSpecification
- Status
- Listed
- Category
- AI
- Services
- clickhouse, db, langfuse-web, langfuse-worker, minio, redis
- Ports
- 3000
- Volumes
- 4
- Memory
- about 4 GiB
- arm64
- image published
- amd64
- image published
- Settings to fill in
- none
- License
- Permissive license, MIT
About Langfuse
Trace, evaluate, and debug LLM applications with prompt management and cost analytics. Runs 6 containers (clickhouse, db, langfuse-web, langfuse-worker, minio, redis), with langfuse-web listening on port 3000; data is kept in named volumes.
What it deploys
Langfuse runs as 6 services in a single Compose app, listening on port 3000, with 4 named volume mounts.
| Service | Image | Tag | Ports | Volumes |
|---|---|---|---|---|
| clickhouse | clickhouse/clickhouse-server | 24.12-alpine | none | langfuse_clickhouse_data -> /var/lib/clickhouse |
| db | postgres | 16-alpine | none | langfuse_db_data -> /var/lib/postgresql/data |
| langfuse-web | langfuse/langfuse | 3.225.11 | 3000 | none |
| langfuse-worker | langfuse/langfuse-worker | 3.225.11 | none | none |
| minio | ghcr.io/coollabsio/minio | RELEASE.2025-10-15T17-29-55Z | none | langfuse_minio_data -> /data |
| redis | redis | 7-alpine | none | langfuse_redis_data -> /data |
Before you install
- CPU architecture. arm64 listed. Every image lists linux/arm64 in its registry manifest. That is a manifest fact, not proof the app runs on Arm.
- Listed. It is in the catalog. Back up the data volume before changing an image tag.
- Host ports. It publishes port 3000 directly on the host, so open them in the VPS firewall only if you want that.
Settings and environment
You do not have to fill in anything. 14 secrets are generated at deploy and stored encrypted. Values are never shown here.
| Service | Variable | Value |
|---|---|---|
| clickhouse | CLICKHOUSE_PASSWORD | Generated at deploy |
| clickhouse | CLICKHOUSE_USER | Preset by the template |
| db | POSTGRES_DB | Preset by the template |
| db | POSTGRES_PASSWORD | Generated at deploy |
| db | POSTGRES_USER | Preset by the template |
| langfuse-web | CLICKHOUSE_CLUSTER_ENABLED | Preset by the template |
| langfuse-web | CLICKHOUSE_MIGRATION_URL | Preset by the template |
| langfuse-web | CLICKHOUSE_PASSWORD | Generated at deploy |
| langfuse-web | CLICKHOUSE_URL | Preset by the template |
| langfuse-web | CLICKHOUSE_USER | Preset by the template |
| langfuse-web | DATABASE_URL | Generated at deploy |
| langfuse-web | ENCRYPTION_KEY | Generated at deploy |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_ACCESS_KEY_ID | Preset by the template |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_BUCKET | Preset by the template |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_ENDPOINT | Preset by the template |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_FORCE_PATH_STYLE | Preset by the template |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_REGION | Preset by the template |
| langfuse-web | LANGFUSE_S3_EVENT_UPLOAD_SECRET_ACCESS_KEY | Generated at deploy |
| langfuse-web | NEXTAUTH_SECRET | Generated at deploy |
| langfuse-web | NEXTAUTH_URL | Preset by the template |
| langfuse-web | REDIS_HOST | Preset by the template |
| langfuse-web | SALT | Generated at deploy |
| langfuse-worker | CLICKHOUSE_CLUSTER_ENABLED | Preset by the template |
| langfuse-worker | CLICKHOUSE_MIGRATION_URL | Preset by the template |
| langfuse-worker | CLICKHOUSE_PASSWORD | Generated at deploy |
| langfuse-worker | CLICKHOUSE_URL | Preset by the template |
| langfuse-worker | CLICKHOUSE_USER | Preset by the template |
| langfuse-worker | DATABASE_URL | Generated at deploy |
| langfuse-worker | ENCRYPTION_KEY | Generated at deploy |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_ACCESS_KEY_ID | Preset by the template |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_BUCKET | Preset by the template |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_ENDPOINT | Preset by the template |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_FORCE_PATH_STYLE | Preset by the template |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_REGION | Preset by the template |
| langfuse-worker | LANGFUSE_S3_EVENT_UPLOAD_SECRET_ACCESS_KEY | Generated at deploy |
| langfuse-worker | REDIS_HOST | Preset by the template |
| langfuse-worker | SALT | Generated at deploy |
| minio | MINIO_ROOT_PASSWORD | Generated at deploy |
| minio | MINIO_ROOT_USER | Preset by the template |
Install Langfuse on a VPS
-
Install Levelrail on a fresh VPS
Run this as root on a Linux server (amd64 or arm64) with ports 80 and 443 reachable. The installer prints a setup link.
curl -fsSL https://levelrail.com/install.sh | sudo sh -
Sign in
Open the setup link, choose a password, then log the CLI in. The CLI install command is
curl -fsSL https://levelrail.com/install-cli.sh | shif the machine does not have it. See Getting started.levelrail-cli login --device -
Read the template
Shows the entry, including its full compose.yaml.
levelrail-cli templates get langfuse -
Deploy it
This deploys the template compose file as an app. The app name defaults to the template id; --name sets another.
levelrail-cli templates deploy langfuse --name my-langfuse
Without a domain the app gets an sslip.io address. You can also deploy it from the dashboard: Apps, New app, Browse templates, then Langfuse.
Docs and upstream
More in AI
- LangflowA visual low-code tool for building and deploying multi-agent and RAG workflows in Python.
- Label StudioA flexible data labeling tool for text, images, audio, and video, with export to common ML formats.
- LibreChatA self-hosted ChatGPT-style interface for many model providers, with agents, search, and multi-user login.
- Kokoro TTSA CPU text-to-speech server with an OpenAI-compatible /v1/audio/speech API, built on the Kokoro model. Apache-2.0 licensed. The image is about 3 GB and there is no login, so keep it on a private network or behind an auth proxy.
- LightRAGGraph based retrieval augmented generation server with a web UI and API. Runs on local file storage; point it at a model provider with the LLM and embedding settings.
- LiteLLM ProxyOne OpenAI-format gateway in front of 100+ model providers, with keys, budgets, and spend tracking.