---
title: "Deploy Langflow on StackBlaze"
description: "Deploy Langflow with managed Postgres, a persistent volume, and a public HTTPS URL. Bring your own model keys. You pay for compute, not per flow run or per seat."
category: "ai, development, utilities"
url: https://www.stackblaze.com/templates/langflow
---

# Deploy Langflow

Deploy Langflow with managed Postgres, a persistent volume, and a public HTTPS URL. Bring your own model keys. You pay for compute, not per flow run or per seat.

**[Deploy Langflow on StackBlaze](https://dashboard.stackblaze.com/new/template/langflow)**

## This stack

- **Langflow**: Web service
- **PostgreSQL**: managed database · TCP:5432

## What you get

- **Visual flow builder**: Drag agents, prompts, models, vector stores, and tools onto a canvas and connect them. Test each flow in the built-in Playground.
- **Any model provider**: Pick OpenAI, Anthropic, Google, a local model, or any provider Langflow supports. Keys are yours and stay on this instance.
- **Every flow is an API**: Call a flow over HTTPS from your app, or expose a project as an MCP server for your agents and editors.
- **Encrypted keys**: Model keys saved as Global Variables are encrypted with a secret key generated at deploy. It is unique to this instance.
- **Managed Postgres**: Flows, users, variables, and message history live in Postgres, not in a SQLite file inside the container.
- **Admin login at deploy**: An admin account is created on first start with a generated password. Auto-login is off, so the URL is not open to anyone.

## After you deploy

1. **Open the public URL**: The hostname is on the app page after deploy. The first start runs database migrations and can take two or three minutes.
2. **Sign in as admin**: Copy LANGFLOW_SUPERUSER_PASSWORD from the app’s Variables in the dashboard and sign in as admin.
3. **Add a model key and start from a template**: Save your provider key in Settings → Global Variables, then pick a starter such as Simple Agent or Vector Store RAG from New Flow.

## How it is wired

- **Postgres holds the workspace**: Flows, users, Global Variables, and message history live in the attached database. Restart the app and everything is still there.
- **Files sit on a volume**: Uploaded files, profile pictures, and Langflow’s own secret_key live on a 5 GB volume mounted at /app/langflow.
- **Keep the secret key**: LANGFLOW_SECRET_KEY encrypts every saved key. It is generated once at deploy. Changing it later makes the saved keys unreadable.
- **Add memory for big flows**: The app starts with 1 GB reserved and can use up to 3 GB. Raise CPU and RAM from the dashboard when flows load large documents or many components.

## Common questions

### How do I sign in after deploy?

Open the public URL and sign in as admin. The password is generated at deploy and shown as LANGFLOW_SUPERUSER_PASSWORD in the app’s Variables in the dashboard.

### Does it include an LLM?

No. Langflow connects to the model provider you choose. Save your API key in Settings → Global Variables, or point a model component at a model you host yourself.

### Where are flows and keys stored?

Flows, users, and Global Variables are in Postgres. Keys are encrypted with LANGFLOW_SECRET_KEY, which is generated at deploy. Uploaded files are on the app volume. Back up the database if the workspace matters.

### Can my app call a flow?

Yes. Each flow has an API endpoint on your public hostname. Create an API key in Langflow settings and call the flow from your code.

### Can I use my own domain?

Yes. Attach a custom domain from the dashboard. The editor and the flow API move with it.

### Can other people use it?

Yes. Sign in as admin and add users in the admin page. New users start inactive until an admin turns them on.

### What does it cost?

You pay for the compute, storage and bandwidth Langflow and Postgres use, not per flow run or per seat. Model usage is billed by your provider.


Source: https://github.com/langflow-ai/langflow