Customer portal
Website
← Knowledge Base

Deploy Flowise on a VPS to build AI workflows visually

Flowise offers a visual way to connect model calls, retrieval and application logic. Start with a simple prompt-to-answer flow for drafting or classification. Add tool execution only after the text-only path is reliable. This example uses the current email/password administrator flow, rather than deprecated basic-username environment variables.

Target: Ubuntu Server 24.04 LTS, x86-64, SSH and a sudo account. Documentation reviewed 15 November 2026. Commands have not been tested end to end on Ubuntu.

Before you start

Plan about 2–4 vCPUs, 4 GB RAM and 15 GB free disk when model inference is external. More document parsing, workers or local inference need more. External providers require your own credentials and usage budget; a flow editor does not include model capacity.

Resource figures are planning estimates for a small evaluation, not performance guarantees or plan specifications. A VPS/VDS does not imply a GPU. Check available memory, storage and resources before ordering.

On a fresh Ubuntu installation, use the packages below. If Docker is already installed from another repository, keep that installation and check compatibility first: do not mix Docker CE and docker.io packages. Use Compose v2; if the package is missing, check that Ubuntu Universe is enabled.

sudo apt update
sudo apt install -y docker.io docker-compose-v2 curl openssl
sudo systemctl enable --now docker
sudo docker compose version

All examples publish ports only on the server’s 127.0.0.1. Access them from your computer through an SSH tunnel; a domain is unnecessary. Do not open these ports to the Internet. An internal container listener on 0.0.0.0 does not change the host’s loopback-only published port.

Each project below is a new installation. Do not rerun secret-generation steps against an existing deployment.

1. Keep state and secrets stable

The current container uses /home/node/.flowise. Generate separate secrets, then save compose.yaml. Keep the environment file: losing the encryption secret can make stored provider credentials unreadable.

umask 077
mkdir -p ~/ai-flowise
cd ~/ai-flowise
sudo install -d -m 700 -o 1000 -g 1000 data
for key in FLOWISE_SECRETKEY_OVERWRITE JWT_AUTH_TOKEN_SECRET JWT_REFRESH_TOKEN_SECRET EXPRESS_SESSION_SECRET TOKEN_HASH_SECRET; do
  printf '%s=%s\n' "$key" "$(openssl rand -hex 32)"
done > .env
services:
  flowise:
    image: flowiseai/flowise:latest
    restart: unless-stopped
    env_file: .env
    environment:
      PORT: "3000"
      APP_URL: http://localhost:3000
      DATABASE_PATH: /home/node/.flowise
      SECRETKEY_PATH: /home/node/.flowise
      BLOB_STORAGE_PATH: /home/node/.flowise/storage
      LOG_PATH: /home/node/.flowise/logs
    ports:
      - "127.0.0.1:3000:3000"
    volumes:
      - ./data:/home/node/.flowise

2. Bootstrap the owner privately

Run the service and connect from your computer. Open http://localhost:3000 to match APP_URL, then create the administrator. If a different application already uses local port 3000, choose another port consistently in the mapping, tunnel and APP_URL.

sudo docker compose up -d
curl --fail-with-body http://127.0.0.1:3000/api/v1/ping
ssh -N -L 127.0.0.1:3000:127.0.0.1:3000 YOUR_USER@YOUR_SERVER_IP

3. Build one small workflow

Create a Chatflow with a chat model and a prompt suited to one narrow task, such as classifying a sample support request. Add the provider credential through the credential store, save and test in the editor. Do not attach company databases or shell tools for this first test. Menu and node names can change between releases; choose the node matching your configured provider.

4. Protect the prediction endpoint

The editor login and a flow’s API access are separate controls. Assign an API key to a flow before exposing its prediction endpoint and call it from your backend, not from browser code that would reveal the key. Keep the editor private. Back up data and .env together; export flow definitions too, but do not assume an export contains every secret or database record. Set provider spending limits before connecting automated triggers.

Updates and recovery

Before upgrading, stop the application for a consistent backup, keep data and secrets off the VPS, and test recovery on a separate instance. Moving image tags in these examples are for initial evaluation; pin a reviewed release tag or digest before ongoing use. Record the current digest with sudo docker image inspect IMAGE_NAME --format '{{json .RepoDigests}}'. Avoid unattended upgrades without compatibility checks, and do not delete persistent volumes with down -v.

If startup fails, inspect sudo docker compose ps, recent logs, free disk and memory. Do not share logs containing keys, prompts or personal data. For directory-permission errors, check the expected UID/GID and ownership instead of granting 777. Set up HTTPS, authentication and request limits before sharing access.

Related deployments