Lesson 4 of 7
Run AI on your laptop
6 min read
You don't always need the cloud. With one free app, a capable AI can run entirely on your laptop — no internet, no account, no bill. What's the catch?
Your machine, your model
Apps like Ollama and LM Studio let you download an open model and run it locally. Everything happens on your computer: your words never leave the machine, it works on a plane, and it costs nothing per question. The trade is size — a laptop can't hold the giant frontier models, so you run a smaller one.
A model's size is really a memory bill. It has to fit in your machine's memory to run well — and the biggest models simply don't fit on a laptop.
Shrink it to fit
There's a clever trick: quantization. It stores the model's numbers at lower precision — like saving a photo at a smaller file size — so it takes far less memory and runs faster. You lose a little sharpness, but a quantized model that fits beats a perfect one that won't load at all.
Quantization trades a sliver of quality for a big drop in memory. It's what makes a real model fit on ordinary hardware.
Local models are smaller, so they're weaker than the cloud's best — more mistakes, less range. Perfect for private drafts, offline notes, and tinkering; reach for the cloud when the job is hard or high-stakes.
The shape of it
- —Ollama and LM Studio run an open model on your own machine — offline, private, free per use.
- —A model must fit in your memory; the biggest ones won't fit on a laptop.
- —Quantization shrinks a model to fit, trading a little quality for a lot of room.
A big model won't load on your laptop — it runs out of memory. What's the usual fix?
Continue in the app
Take the whole Which AI should I use? course — tracked
Get your personalized path, progress and streaks in the app — this lesson and every next one, in order.