Boltjar and n8n
Boltjar owes n8n its other core idea: workflows that start from triggers and keep running, instead of scripts you run by hand. The two tools point that idea at different jobs.
What each is for
Section titled “What each is for”n8n is a workflow automation platform. Its strength is integrations: hundreds of ready-made nodes for business apps and services, credentials handled for each, and a visual canvas you can mix with code. It has AI agent features, runs self-hosted or in n8n’s cloud, and is distributed under its own fair-code Sustainable Use License.
Boltjar is a visual node editor for always-on AI systems that you run on your own machine. Its nodes are small and typed: a chat model, a template, a vector search, a speech model, a SQLite query, an HTTP call. You compose a system from those pieces, and every wire has a type the editor checks. It is licensed AGPL-3.0 with a node pack exception, and it has no cloud version.
Side by side
Section titled “Side by side”| n8n | Boltjar | |
|---|---|---|
| Built for | connecting apps and services | composing AI systems from small nodes |
| Integrations | hundreds of app nodes | generic nodes: HTTP Request, Webhook, SQLite, files |
| Where it runs | self-hosted, or n8n’s cloud | on your machine, loopback by default |
| Wires | items flow from node to node | typed pipes: events push, data is pulled |
| AI | agent and model nodes | LLM, tools, speech, embeddings, vectors as nodes |
| Local models | yes | Ollama, first-class |
| Custom nodes | community nodes in TypeScript | node packs in Python |
Which one to use
Section titled “Which one to use”Use n8n when the job is moving data between business apps and the integration already exists. Use Boltjar when the job is an AI system you want to shape piece by piece on your own machine: a local assistant with memory, an agent whose tools are nodes, a voice loop, retrieval over your own notes. They also meet in the middle: an n8n workflow can call a Boltjar Webhook, and a Boltjar HTTP Request can call n8n.