This guide covers downloading, installing, and configuring Flowise for building AI agents and LLM workflows.
Ensure you have:
# Install globally
npm install -g flowise
# Start Flowise
npx flowise start
Open your browser to:
# Install globally
npm install -g flowise
# Start Flowise
npx flowise start
# Access at http://localhost:3000
Step 1: Clone repository
git clone https://github.com/FlowiseAI/Flowise.git
cd Flowise/docker
Step 2: Configure environment
cp .env.example .env
# Edit .env with your settings
Step 3: Start services
docker compose up -d
# Clone repository
git clone https://github.com/FlowiseAI/Flowise.git
cd Flowise
# Build image
docker build --no-cache -t flowise .
# Run container
docker run -d --name flowise -p 3000:3000 flowise
# Clone repository
git clone https://github.com/FlowiseAI/Flowise.git
cd Flowise
# Install dependencies
pnpm install
# Build
pnpm build
# Start
pnpm start
| Component | Requirement |
|---|---|
| Node.js | 18.15.0+ |
| RAM | 4 GB |
| Disk | 10 GB |
| Package Manager | PNPM >= 10.26.0 |
| Component | Requirement |
|---|---|
| Node.js | 20.x |
| RAM | 8+ GB |
| Disk | 20+ GB |
| Package Manager | Latest PNPM |
Create .env file:
# Port
PORT=3000
# Database
DATABASE_TYPE=sqlite
DATABASE_PATH=/path/to/database
# API Keys
FLOWISE_API_KEY=your-api-key
# Logging
LOG_LEVEL=info
# CORS
CORS_ORIGINS=*
SQLite (Default):
DATABASE_TYPE=sqlite
DATABASE_PATH=/path/to/database.sqlite
PostgreSQL:
DATABASE_TYPE=postgres
DATABASE_HOST=localhost
DATABASE_PORT=5432
DATABASE_NAME=flowise
DATABASE_USER=flowise
DATABASE_PASSWORD=your-password
MySQL:
DATABASE_TYPE=mysql
DATABASE_HOST=localhost
DATABASE_PORT=3306
DATABASE_NAME=flowise
DATABASE_USER=flowise
DATABASE_PASSWORD=your-password
# Disable telemetry
DISABLE_TELEMETRY=true
# Enable authentication
BASIC_AUTH_USERNAME=admin
BASIC_AUTH_PASSWORD=secure-password
# Configure storage
STORAGE_TYPE=local
STORAGE_PATH=/path/to/storage
# Redis for scaling
REDIS_HOST=localhost
REDIS_PORT=6379
REDIS_PASSWORD=your-redis-password
pip install flowise
from flowise import Flowise, PredictionData
# Initialize client
client = Flowise()
# Create prediction
completion = client.create_prediction(
PredictionData(
chatflowId="your-flow-id",
question="What is the capital of France?",
streaming=False
)
)
for response in completion:
print(response)
from flowise import Flowise, PredictionData
client = Flowise()
completion = client.create_prediction(
PredictionData(
chatflowId="your-flow-id",
question="Tell me a joke!",
streaming=True
)
)
for chunk in completion:
print(chunk, end="", flush=True)
| Endpoint | Method | Purpose |
|---|---|---|
/api/v1/prediction/{chatflowId} |
POST | Create prediction |
/api/v1/chatflows |
GET/POST | List/create flows |
/api/v1/nodes |
GET | List available nodes |
curl http://localhost:3000/api/v1/prediction/{chatflowId} \
-H "Content-Type: application/json" \
-d '{
"question": "What is the capital of France?"
}'
# Clear npm cache
npm cache clean --force
# Reinstall
npm install -g flowise
# Change port
export PORT=3001
npx flowise start
# Reset SQLite database
rm database.sqlite
# Restart Flowise
npx flowise start
# Increase Node.js memory
export NODE_OPTIONS="--max-old-space-size=4096"
npx flowise start
# Update globally
npm update -g flowise
# Restart Flowise
npx flowise start
# Pull latest image
docker pull flowise/flowise:latest
# Restart container
docker restart flowise
Any questions?
Feel free to contact us. Find all contact information on our contact page.