b8978

b8978 is a release of llama.cpp, a project designed to enable LLM inference with minimal setup and state-of-the-art performance on a wide range of hardware. The llama.cpp project uses sequential build numbers like b8978 rather than traditional semantic versioning, with multiple releases published in a single day as features and improvements are merged. This specific build incorporates updates and optimizations to the underlying C/C++ implementation for running large language models efficiently.

Source: llama.cpp Releases | 2026-04-29