llama.cpp b9981 lands with cpp-httplib 0.50.1 update
Latest release brings a critical HTTP library bump to 0.50.1.
ggml-org has pushed llama.cpp release b9981, a maintenance tag that updates the vendored cpp-httplib dependency to version 0.50.1. The change (PR #25576) addresses several bug fixes and security improvements in the underlying HTTP library used by llama.cpp's built-in server and API endpoints. The release is GPG-signed and verified, ensuring authenticity for production deployments.
Beyond the library update, b9981 offers an extensive matrix of pre-compiled binaries. macOS builds cover Apple Silicon (with optional KleidiAI acceleration) and Intel x64, plus an iOS XCFramework. Linux users get builds for x64, arm64, and s390x CPUs, with Vulkan, ROCm 7.2, OpenVINO, and SYCL FP32/16 backends. Windows binaries include CPU-only, arm64, OpenCL Adreno, CUDA 12/13, Vulkan, OpenVINO, SYCL, and HIP. Android arm64 and openEuler variants are also included. This release mirrors the project's commitment to making local LLM inference accessible across platforms.
- Updated internal cpp-httplib from previous version to 0.50.1 (PR #25576) for security and stability
- Release provides pre-built binaries for macOS, Linux, Windows, Android, iOS, and openEuler with multiple GPU backends
- Commit is GPG-signed and verified, enabling secure production use
Why It Matters
Keeps llama.cpp's HTTP server reliable for developers running local AI models in production.