Comparison
Also self-hostable, also source-available
Not a hosting argument — you already run your own infrastructure. It is about what arrives beside the workflow engine: assistants, cited retrieval, meeting notes and a PII layer, on one permission model and one audit trail. And n8n is very good; we integrate with it, which should tell you we do not think you have to choose.
Try the builder All comparisonsWhere the two differ
That is the whole comparison, and which side of it you want depends entirely on what you are trying to build.
The node catalogue. n8n has spent years building integrations and a community that builds more, and if your workflow needs a connector to something specific, the odds are far better there than here. It is also a more focused product: if a workflow engine is all you want, everything else on this page is weight you are carrying for nothing. Engineers who like it tend to like it a lot, and that is worth taking seriously.
Assistants with memory and cited answers from your own documents, meeting transcription and notes, and a chat surface — all sharing the same permission model and audit trail as the automations. Not a language-model node in a flow, but a workspace the flows live inside.
Install the detector and personal data is tokenised before any prompt reaches a model provider, on your hardware, restored in the reply. It covers the chat, the assistants and the automations at once rather than being something each flow has to remember.
Say what you want in a sentence and watch the canvas assemble step by step, drag twenty-seven step types together by hand, or press "find repeating work" and let Bee Flow read ninety days of your own tool activity and propose automations for patterns it can actually see. The third one is the interesting one: it is the difference between a tool that builds what you ask for and one that tells you what is worth building.
Where a node does not exist yet, MCP fills the gap rather than a feature request: any Model Context Protocol server plugs in over stdio or Streamable HTTP, and its tools become available to assistants and automations alike with the same permissions and the same audit trail as a built-in integration. It is a different shape of answer to the problem a large node catalogue solves — worth weighing against each other rather than counting.
How to decide
If workflows are the thing you are building and your team is comfortable in a node editor, n8n is a focused, mature answer and you should probably use it. If the workflows exist to serve people who mostly want to ask questions of their own documents and get work done in chat, then the workspace matters more than the engine.
The first automation gets built by whoever is most technical. The tenth gets built by someone in finance who has never seen a node editor, or it does not get built. Describing a workflow in a sentence and correcting what appears is a lower floor than a blank canvas — that is the whole argument for the chat builder.
Every Monday at 07:30, summarise what changed and mail it to the team
If it does, and if a language model is involved, the question becomes who sees the names and account numbers on the way out. A detector running on your own hardware, applied to everything rather than configured per flow, is the part that is hard to bolt on afterwards.
21 categories · on-premise · tokenise and restore
n8n is a Bee Flow integration. If you already have flows you are happy with, keep them: Bee Flow can trigger them and read what they return, so an assistant can call an n8n workflow as an ordinary tool. That is a genuine answer, not a diplomatic one — rebuilding a working automation to move it between platforms is rarely the best use of anybody’s month.
Both projects are source-available rather than OSI open source, and both restrict reselling the product as a hosted service. We are not going to pretend there is a licensing gap between us where there is not one. Bee Flow is under the Sustainable Use License v1.0 and we are working toward AGPL-3.0-or-later; read ours, read theirs, and judge them side by side.
Questions
No — there is no importer, and a half-working one would cost you more time than rebuilding. The better path is usually to leave the flows where they are and connect them: n8n is a built-in integration, so an assistant or automation can trigger an existing workflow and use what it returns.
Not as a node catalogue, no — and the count is the wrong axis to compare on. We ship 44 built-in integrations — Google Workspace, Microsoft 365, Nextcloud, ITSM, payroll, Dutch legal sources — and then the ceiling comes off, because Bee Flow speaks the Model Context Protocol. Any MCP server becomes a set of tools your assistants and automations can call, over stdio or Streamable HTTP: the public ones, the ones your own vendors publish, and the one you write in an afternoon for the line-of-business system nobody has ever built a connector for. When you self-host, installing those is your decision and nobody has to add your system to a catalogue first — the install surface is an enterprise feature, and once a server is installed, using it is an ordinary integration with the same permissions and audit trail as the rest. Where a large catalogue genuinely wins is the long tail of SaaS you want working in ten minutes without writing anything, so check the specific apps you depend on before you switch.
For control flow, close: 27 step types with conditions, switches, loops, parallel branches, filters, deduplication, waits, HTTP calls and code, plus approval gates and dry runs. Where we are ahead is the AI side — building by description, and automations proposed from observed activity. Where n8n is ahead is breadth of nodes.
Because they answer different questions. Teams commonly keep n8n for the integration-heavy plumbing they already trust, and use Bee Flow for the AI workspace — assistants over their documents, meeting notes, the privacy layer — with Bee Flow calling into n8n where a flow already exists.
Source-available, and we would rather be precise than borrow the word. The server and frontend are published under the Sustainable Use License v1.0: read it, modify it, run it internally without paying — what you may not do is resell Bee Flow as your own hosted service. It is not an OSI-approved open-source licence. We are working toward AGPL-3.0-or-later; today the Nextcloud connector is the only part that carries it.
The real automation builder runs in your browser on this site, with sample data and nothing saved. Ten minutes with it will tell you more than this page can.
Try the builder Run the whole stack