Langflow
Visual flow builder for AI agents in Python — open source, MIT-licensed, and infinitely hackable.
Langflow is an MIT-licensed visual builder for AI workflows and agents, beloved by Python teams for one defining trait: every node on the canvas is Python you can open and rewrite. Drag components for models, prompts, vector stores, tools, and agents; connect them into RAG pipelines or multi-agent systems; test in the built-in playground; then serve every flow as an API or export it as JSON for version control. The component library spans all major LLM providers and vector databases, with MCP support letting flows act as tools for other agents. Acquired by DataStax (now under IBM), it ships both as a pip-installable local app and a managed cloud with a generous free tier. Compared to Flowise it's the Python-native twin; compared to Dify it's a flexible construction kit rather than a full platform — the choice of builders who eventually want to read and own the code underneath.
Pros
- MIT license — zero restrictions on any use
- Python-editable components beat closed node systems
- Strong free cloud tier for instant starts
Cons
- Less production tooling (observability, evals) than platforms like Dify
- Heavier resource footprint locally on big flows
