WebGPU

llama.cpp adds local LLM inference in C/C++

ggml-org’s llama.cpp keeps expanding local LLM support with OpenAI-compatible serving, browser WebGPU, and broad hardware backends.