Free AI Update Makes Your Apple M2 Mac Faster
New free tweak speeds up on-device AI on M2 Macs.
If you've ever used ChatGPT, you've used a cloud AI — your request goes to a server, and the answer comes back. But there's a growing movement to run AI directly on your own device. llama.cpp is one of the most popular free tools for this, letting developers build AI apps that work offline and keep your data private.
This latest release (version b10710) is all about speed on Apple's M2 chip. Specifically, it improves how the software handles certain types of compressed AI models — think of compression like a zip file for AI. Compressed models are smaller and quicker, but need clever software to run well. This update adds optimizations for a few more of those formats, making them run noticeably faster on M2-powered Macs and iPads.
What does that mean for you? If you use an AI app that runs on your device — like an offline writing assistant or a photo organizer — this update could make it feel snappier and drain less battery. It also lowers the barrier for developers to build new on-device AI features, which often means better privacy for everyday users.
The catch: this is a technical update for developers, not a button you can click in an app. If you're not a programmer, you'll likely only notice the benefits once your favorite apps update to use the latest version. But each small improvement like this adds up to a future where AI works smoothly right on your own devices.
- llama.cpp is a free, open-source tool that lets AI run on your own device instead of the cloud.
- This update speeds up AI models on Apple M2 chips — especially for compressed models that save space and power.
- Faster on-device AI means faster apps, better privacy, and no need for an internet connection.
Why It Matters
Your future AI apps will run faster and keep your data private on Apple devices.