Can a 16GB M6 Mac Mini Really Handle Local AI?

Can a 16GB M6 Mac mini actually do useful work with local AI? In this video, I put Gemma 4 12B through five everyday tasks to find out. We test whether it can: • Read an invoice image and extract the details • Build a small invoice-processing app • Read a PDF and answer questions about it • Search the web for information • Organize a messy folder of invoices I use LM Studio and Pi to explore what happens when you give a small local model access to practical tools. You’ll see what worked, what struggled, and where the setup got in the way. 📺 ALL my local AI videos for the 16gb Mac Mini: htt