Compare
n8n vs the local stack.
n8n is the workflow engine; it moves data between tools. But the "AI" in your automation still has to run somewhere. Run it locally and n8n becomes the glue for a stack you own, not a bill you pay.
What each layer does
n8n is the engine, the local stack is the car
n8n (the engine)
- Orchestrates steps: trigger → think → write → validate
- Runs locally (Docker), no n8n cloud fee
- Connects to any API or local service
The local stack (the car)
- Ollama: your models, on your GPU
- ComfyUI: local image generation
- Qdrant: private vector search
- SearXNG: self-hosted search
Why it matters
No per-step SaaS tax
A cloud automation charges you every time a step runs, every time a model answers, every time an image renders. Pin n8n to a local stack and the only recurring cost is electricity. The n8n Freelance Ops kit ships five production workflows you import and run the same day.