Multi-Model Advisor is an MCP server that orchestrates queries across several Ollama models, each assigned a distinct persona, to present diverse AI perspectives on a single question. It exposes two MCP tools: list-available-models and query-models, enabling users to discover installed models and run multi-model queries. The server aggregates the responses from each model and Claude or Desktop to synthesize a comprehensive answer, effectively forming a 'council of advisors.' Configuration is done via an environment file (.env) where you specify server name and version, Ollama API URL, and default models. For each model you can set a separate system prompt to shape its role. Example settings include gemma3:1b, llama3.2:1b, and deepseek-r1:1.5b with prompts to encourage creativity, empathy, or analytical reasoning. The server is designed to run with Node.js (Node 16+), Ollama installed, and Claude for Desktop integration. It supports plugging into Claude's desktop tooling for an enhanced advisory experience and includes troubleshooting guidance for Ollama connectivity, model availability, and memory constraints.