Developer Documentation

Installation

npm install -g flowise
npx flowise start

Flowise runs as a local server (default: http://localhost:3000) with a full visual canvas UI.

API Usage

All deployed flows expose a prediction endpoint:

import requests

url = "/api/v1/prediction/:id"

def query(payload):
    response = requests.post(url, json=payload)
    return response.json()

output = query({"question": "hello!"})

SDKs are available for TypeScript and Python.

Key Technical Features

• Multi-agent orchestration via Agentflow with distributed task coordination
• RAG support across file types: TXT, PDF, RTF, DOC, HTML, CSS, JS, JSON, XML, CSV, MD, SQL
• Observability: Full execution traces, Prometheus metrics export, OpenTelemetry integration
• Embedded Chat Widget: Drop into any web app with a JS snippet
• 100+ LLMs & embeddings: OpenAI, Anthropic, Mistral, Llama, AWS Bedrock, and more
• Vector DB integrations: Pinecone, Milvus, Chroma, Weaviate, and others
• Horizontal scaling: Message queue + worker architecture for production workloads
• On-prem & cloud: Docker-ready, supports self-hosted and managed cloud deployments