🛠️ Instructions

FastFlowLM (FLM) is a deeply optimized runtime for local LLM inference on AMD NPUs —
ultra-fast, power-efficient, and 100% offline.

Its user interface and workflow are similar to Ollama, but purpose-built for AMD’s XDNA architecture.

This section will walk you through how to use FastFlowLM with examples.

📚 Sections

• System Command and CLI Mode
• Server Mode
• Server Basics
• API / Client Usage
• Open WebUI
• LangChain RAG
• LangChain Web Search
• Obsidian
• Microsoft AI Toolkit