The Free Tool That Runs AI Chatbots on Your Own Laptop Just Updated

WHAT HAPPENED: A popular free software project called llama.cpp — think of it as an engine that runs AI chatbots on your own device — published a new numbered version, b11076. The actual code change was tiny: developers removed an outdated comment in the test files. Most of the release is freshly compiled builds, meaning ready-to-install files for Windows, Mac, Linux, Android, and iOS. The project also has a website at llama.app.

WHY YOU CARE: Almost every AI chatbot you use today runs on someone else's computer. You type a message, it travels to a data center, gets processed, and comes back. That means you need internet, you often pay a subscription, and your words pass through a company's servers. llama.cpp does the opposite — it runs the AI on hardware you already own. For a professional worried about confidentiality, that's a real difference. Contract details, medical notes, or unpublished drafts can stay on your machine.

WHAT IT FEELS LIKE IN PRACTICE: Imagine a small chatbot app on your laptop that keeps working on an airplane with no Wi-Fi. That's the promise. Setup is not quite point-and-click yet — you generally download a model file (often several gigabytes) and run a command — but the community has made it steadily easier, and dozens of free apps are built on top of it.

THE CATCH: This particular release is a pre-release build, meaning it's a snapshot for testing rather than a polished consumer product. There's no new feature to get excited about. And while your data stays private, the AI you run this way is usually smaller and less capable than the giant cloud models — good enough for summarizing, drafting, and translating, but not for the hardest questions. Also, more capable models need beefier computers.