llama.cpp b9748 adds verbose field to server schema

The open-source llama.cpp repository, known for running large language models locally on consumer hardware, has pushed release b9748. This minor but impactful update introduces a 'verbose' field to the server's OpenAPI-compatible schema. Users can now enable detailed logging directly from API requests, aiding in debugging model interactions and server performance. The change was merged via pull request #24864 and is signed with a verified GPG key.

The release packages binaries for nearly every major platform: macOS Apple Silicon (with optional KleidiAI acceleration) and Intel, Linux distributions (Ubuntu x64/arm64 with CPU, Vulkan, ROCm 7.2, OpenVINO, SYCL), Windows (x64/arm64 with CPU, CUDA 12/13, Vulkan, OpenCL Adreno), Android arm64, and iOS XCFramework. This wide compatibility ensures developers and hobbyists can quickly update their local LLM deployments. The project continues to lead in on-device inference with over 118,000 GitHub stars and active community contributions.