New AI Update Lets You Run LLMs on Your Phone
This free tool lets you run advanced AI models offline on almost any device...
Deep Dive
According to the llama.cpp release page, a new pre-release version b10604 is available. The release includes changes for Deepseek 4, such as sm tensor adjustments and model saving support. It also offers builds for macOS, iOS, Linux, Android, Windows, and openEuler across various hardware backends like Vulkan, CUDA, ROCm, OpenVINO, and SYCL.
Key Points
- A free tool called llama.cpp now lets you run AI models on your phone or old laptop—no internet or fancy computer needed.
- This could save money, protect privacy, and make AI more accessible to small businesses and students.
- Setting it up requires some tech skills, but it’s a big step toward portable, offline AI.
Why It Matters
This update could make advanced AI tools portable and private—imagine running a chatbot on your phone without relying on the cloud or big tech.