🔐 Private Local Ollama + Dynamic LLM Router

This n8n workflow enables fully private, self-hosted AI interactions using Ollama. It analyzes incoming prompts to classify tasks—text reasoning, coding, or image analysis—and dynamically routes them to specialized local models like qwq, llama3.2, phi4 for text; qwen2.5-coder for programming; and granite3.2-vision or llama3.2-vision for visuals. Conversation memory maintains context across interactions.

Key benefits include complete data privacy (no cloud APIs), optimal model selection for superior results, and seamless scalability for multiple local LLMs. Ideal for developers, enterprises handling sensitive data, or AI enthusiasts avoiding vendor lock-in.

Use cases: Secure internal chatbots, local code review assistants, image description tools, or research prototypes. Setup requires Ollama installation, model pulls, and n8n credential config—then interact via chat interface for instant, intelligent routing.