On-Device & Edge AI

on-device AI course, edge AI development, Apple Foundation Models tutorial, Gemini Nano Android, MLX Apple Silicon, run LLM locally, on-device LLM for iOS, model quantization tutorial, local AI inference, edge AI deployment for engineers

advanced

en

• Embed a ~3B Apple Foundation Models LLM in an iOS app using guided generation and tool calling
• Ship on-device AI to Android with AICore and the ML Kit GenAI APIs
• Run open models locally on Apple Silicon with MLX, Ollama, and quantization
• Decide between on-device inference and a cloud model like Claude for a given task
• Pick the right edge platform across Apple, Android, and Mac for a use case
• Reason about the throughput and limits of a single-user local inference node

• Url: https://aiacademy.anthropos.work/chapters/on-device-edge-ai-intro/ · On-Device & Edge AI: Start Here · Position: 1 · A 12-minute orientation to the On-Device & Edge AI skill path — the three platforms where your model can run without a cloud round-trip, and how to pick yours
• Url: https://aiacademy.anthropos.work/chapters/apple-foundation-models/ · Apple Foundation Models for Swift Developers · Position: 2 · Ship a capable ~3B LLM inside your iOS app — guided generation, tool calling, and when to still reach for Claude
• Url: https://aiacademy.anthropos.work/chapters/gemini-nano-android/ · Gemini Nano and AICore on Android · Position: 3 · Ship on-device AI to 140M+ Android devices — AICore as a shared system service, ML Kit GenAI APIs, and Gemma 4 agentic intelligence
• Url: https://aiacademy.anthropos.work/chapters/mlx-local-inference/ · MLX in Practice: Local Inference on Apple Silicon · Position: 4 · Run arbitrary open models on your Mac — unified memory, JANG quantization, Ollama + MLX, and the honest limits of a single-user inference node

• Embed a ~3B Apple Foundation Models LLM in an iOS app using guided generation and tool calling
• Ship on-device AI to Android with AICore and the ML Kit GenAI APIs
• Run open models locally on Apple Silicon with MLX, Ollama, and quantization
• Decide between on-device inference and a cloud model like Claude for a given task
• Pick the right edge platform across Apple, Android, and Mac for a use case
• Reason about the throughput and limits of a single-user local inference node

It is built for software and AI engineers, especially iOS, Android, and Mac developers who want to add private, low-latency AI features that work without a server. The level is advanced and assumes you already ship production app code.

Course Mode: online · Course Workload: PT78M · Mode: online