ggerganov/llama.cpp Released Version b12451 with UI Restructuring
Llama.cpp release b12451 restructures the built-in web UI by moving it to a dedicated folder, ui/apps/chat, and renaming it to Llama Chat. This change serves as the first step in separating the UI into a core library (ui/lib) and individual applications built on top of it. This architectural cleanup improves the codebase maintainability and modularity of llama.cpp's front-end components. It lays the groundwork for better organization as the project continues to scale its tooling and web interfaces. The internal code paths, CMake rules, Dockerfiles, Nix packages, and documentation references have all been updated to point to the new ui/apps/chat path. Meanwhile, external-facing compatibility layers, such as deprecated flags and environment variable names, remain unchanged to prevent breaking existing user setups.
## BACKGROUND
llama.cpp is a widely used open-source software library that performs efficient inference on large language models like Meta's Llama. Co-developed alongside the GGML tensor library, it has become a foundational standard for running AI models locally on consumer hardware.