b9333

Release b9333 of llama.cpp adds support for Apple device IDs in the Metal framework for GPU acceleration on macOS and iOS . The build includes compiled binaries for multiple platforms including macOS, Linux, Windows, and Android with various hardware acceleration options (CPU, Metal, Vulkan, CUDA, ROCm, and others). This is part of the llama.cpp project's continuous rapid development cycle, delivering incremental improvements and hardware compatibility enhancements to its C/C++ large language model inference engine.

Source: llama.cpp Releases | 2026-05-26