Smart Local Model Routing for Privacy-Focused AI Chat

This n8n workflow automates intelligent routing of user prompts to appropriate locally hosted large language models (LLMs) using Ollama. When a chat message is received via webhook, the workflow analyzes the user input through an LLM router node, which classifies the task and selects the best-suited Ollama model based on predefined rules and capabilities. Depending on the classification, it dynamically assigns either a text-only, code, vision, or combined model for processing. Conversations are maintained with memory buffers to ensure context-aware responses. Designed for privacy-conscious users and AI enthusiasts, this setup allows seamless, local AI model management without external data transmission, making it ideal for secure AI deployment and customized model utilization.