Langflow v1 AI Agent Builder

Langflow is an open-source visual builder for AI agents, RAG pipelines and MCP servers. This template runs Langflow 1.12.5 with Postgres, Redis and a persistent volume, with generated admin credentials and two Gunicorn workers.

What you get

  • langflowai/langflow 1.12.5 serving the editor, REST API and MCP endpoints on port 7860
  • Postgres 18 for flows, users, messages and encrypted variables instead of SQLite
  • Redis 8.2 as the shared build and job queue and for rate-limit counters, so multiple workers stay consistent
  • A volume at /app/langflow for uploaded files, knowledge bases and logs
  • A generated admin password and encryption key; new sign-ups stay inactive until an admin activates them

Good for

  • Prototyping LLM agents with tools such as web search and APIs without writing a backend
  • Building retrieval-augmented chat over your documents and exposing it as an API
  • Publishing flows as MCP servers for Claude Desktop, Cursor and other MCP clients
  • Giving a team a shared, login-protected workspace for AI workflows

Services in this template

  • Redis · image redis:8.2
  • Langflow · image langflowai/langflow:1.12.5
  • Postgres · image ghcr.io/railwayapp-templates/postgres-ssl:18

Questions people ask

How do I log in?

Open the public URL and sign in as admin with LANGFLOW_SUPERUSER_PASSWORD from the Langflow service variables. The first deploy can take a few minutes while migrations and starter projects load.

Where do I put model provider keys?

Under Settings, Global Variables. They are encrypted with LANGFLOW_SECRET_KEY, which you must never change after the first boot.

How do I add more workers?

Raise LANGFLOW_WORKERS, about one per vCPU at roughly 0.6 to 1 GB RAM each according to the template notes, and give the service more resources. Keep LANGFLOW_WORKERS times the pool size and max overflow below about 90 database connections.

Can I embed the chat widget on my site?

Yes. Add each website origin that embeds Langflow to LANGFLOW_CORS_ORIGINS. For a custom domain, also update LANGFLOW_CORS_ORIGINS and LANGFLOW_MCP_BASE_URL.

How do I upgrade?

Back up Postgres, change the image tag on the Langflow service and redeploy; Langflow migrates the database on start. Do not roll back to an older version against a migrated database without restoring the backup. The NOTICE file lists Langflow as MIT licensed.

Full template README

Deploy and Host Langflow with Railway

Langflow is an open-source visual builder for AI agents, RAG pipelines and MCP servers. You connect components on a canvas, test them in a playground and serve every flow as an API or MCP tool. This community template runs a pinned Langflow release with Postgres, Redis and persistent storage.

About Hosting Langflow

Langflow is a Python application that serves its editor, REST API and MCP endpoints from one process. For a team deployment it needs a real database instead of SQLite, a persistent configuration directory for uploaded files, knowledge bases and logs, a stable encryption key for stored credentials, and a login-protected admin account. Running more than one worker also requires a shared job queue, otherwise editor builds can land on the wrong worker. This template provides all of that on Railway: Postgres for flows and users, Redis for the job queue and rate limits, a volume for the config directory, and generated admin credentials.

Common Use Cases

  • Prototype and ship LLM agents with tools (web search, APIs, code) without writing a backend
  • Build retrieval-augmented chat over your documents and expose it as an API
  • Publish flows as MCP servers for Claude Desktop, Cursor and other MCP clients
  • Give a team a shared, login-protected workspace for AI workflows and prompt experiments
  • Embed a chat widget backed by a Langflow flow in your website or internal tools

Dependencies for Langflow Hosting

  • Langflow (langflowai/langflow, default image profile)
  • PostgreSQL 15 or newer (Railway Postgres 18)
  • Redis (job queue and rate-limit storage for multiple workers)
  • A persistent volume for LANGFLOW_CONFIG_DIR

Deployment Dependencies

Implementation Details

Service Image Role Public
Langflow langflowai/langflow:1.12.5 Editor, API, MCP server; 2 Gunicorn workers; volume at /app/langflow yes (port 7860)
Postgres Railway Postgres 18 Flows, users, messages, encrypted variables no
Redis Railway Redis 8.2 Shared build/job queue (DB 1) and rate-limit counters (DB 2) no

First login

  1. Wait for the first deploy to pass its healthcheck (/health_check); migrations and starter projects can take a few minutes.
  2. Open the public URL and sign in as admin with the password in the Langflow service variable LANGFLOW_SUPERUSER_PASSWORD.
  3. Add your model provider keys under Settings → Global Variables. They are encrypted with LANGFLOW_SECRET_KEY; never change that key after first boot.
  4. New users who sign up stay inactive until an admin activates them (LANGFLOW_NEW_USER_IS_ACTIVE=false).

Scaling workers: raise LANGFLOW_WORKERS (about one per vCPU, 0.6–1 GB RAM each) and give the service more resources. Redis keeps builds consistent across workers. Keep LANGFLOW_WORKERS × (pool_size + max_overflow) from LANGFLOW_DB_CONNECTION_SETTINGS below about 90 connections.

Embedding the chat widget: add each website origin that embeds Langflow to LANGFLOW_CORS_ORIGINS (comma-separated).

Custom domain: add it to the Langflow service, then update LANGFLOW_CORS_ORIGINS and LANGFLOW_MCP_BASE_URL to the new origin.

Pinning and upgrading: the image is pinned to 1.12.5. To upgrade, back up Postgres, change the tag on the Langflow service and redeploy; Langflow migrates the database on start. Do not roll back to an older version against a migrated database without restoring the backup. To change the component inventory, switch between langflowai/langflow:base-{ver}, langflowai/langflow:{ver} and langflowai/langflow-all:{ver}.

Why Deploy Langflow on Railway?

Railway gives Langflow a public HTTPS URL, managed Postgres and Redis on a private network, and a persistent volume in a single project, so your flows, files and credentials survive redeploys. Scaling is a matter of raising workers and resources, and upgrades are a tag change with one-click rollback. This is a community template and is not affiliated with the Langflow project or its maintainers.

Guides for this template

View Langflow v1 AI Agent Builder on railway.com

Similar templates

TemplateCategoryDeploymentsDeploys
9Router
Self-hosted AI router for Claude, Gemini, Cursor & OpenAI tools.
AI/ML9,734
flowise-railway
Flowise - low-code LLM apps builder
AI/ML8,512
LibreChat
Open-source ChatGPT clone: multi-model, agents, search, multi-user auth
AI/ML6,762
Hermes Agent by Nous Research
Template for Hermes Agent by Nous Research
AI/ML6,413
OpenClaw
[Homebrew support] The AI that actually does things
AI/ML5,167
Hermes Agent
Self Host Hermes Agent. Autonomous AI agent w/ self-improving, tools & more
AI/ML4,092