Flowise

Open SourceOtherSelf-hostable

Build LLM apps and AI agents visually — chain models, RAG pipelines, tools and multi-agent systems on a canvas, then ship them as APIs, chatbots or embedded widgets.

Open source alternative to

n8n

Zapier

Visit Flowise View source

About Flowise

Flowise is what you use when the automation you're building is really an AI application. Drag chat models, prompts, vector stores, document loaders, memory and tools onto a canvas, wire them together, and you have a working RAG chatbot or agent — exposed instantly as an API endpoint or an embeddable chat widget.

Where n8n added AI nodes to a general automation tool, Flowise is AI-native: agent orchestration with supervisor/worker patterns, streaming responses, conversation memory, human-in-the-loop steps and evaluations are the core product, not an add-on. Under the hood it builds on the LangChain/LlamaIndex ecosystems, so new model and retrieval techniques land fast.

At 54k+ stars it's one of the fastest-growing projects in open source. Self-host it with Docker to keep prompts and data private, or use Flowise Cloud — either way, non-ML engineers can ship real LLM features in an afternoon.

Key features

Visual LLM orchestration — Chain models, prompts, retrievers and tools on a canvas — see your whole AI pipeline at a glance.
RAG out of the box — Document loaders, chunkers, embeddings and vector stores (Pinecone, Qdrant, pgvector…) built in.
Agents & multi-agent flows — Tool-using agents, supervisor/worker teams, memory and human-in-the-loop approvals.
Ship as API or widget — Every flow becomes a REST endpoint, streaming chat API or embeddable web widget instantly.

How to install Flowise

The honest read

What we like

Fastest visual route from idea to working LLM app
Full RAG stack included — no glue code
Model-agnostic: OpenAI, Anthropic, Gemini, Ollama, local
Apache-2.0 with a massive, fast-moving community
Self-host keeps prompts and data private

Worth knowing

Narrow outside AI — not a general automation tool
Complex flows can outgrow the visual paradigm
Fast-moving upstream libs mean occasional breakage
Production hardening (scaling, evals) takes extra work

FAQ

Is Flowise free?

Yes — Apache-2.0 licensed and free to self-host without limits. Flowise Cloud offers hosted plans if you'd rather not run it yourself.

Flowise vs n8n — which one for AI workflows?

If AI is one step in a broader business workflow, n8n's AI nodes may be enough. If the AI is the product — a RAG chatbot, an agent, a copilot — Flowise's AI-native tooling (vector stores, memory, agent patterns, chat widgets) is far deeper.

Does Flowise work with local models?

Yes — it connects to Ollama, LocalAI, vLLM and other local runtimes, so you can run fully private pipelines with no external API calls.

Can I embed a Flowise chatbot on my site?

Yes — every chatflow ships with an embeddable widget (a script tag) plus REST and streaming APIs for custom frontends.