The llama.cpp project released version b9956, which includes a fix for the command-line interface to prevent crashes when an incorrect server base URL is provided. This change ensures graceful exit by catching exceptions rather than failing abruptly.
- macOS/iOS: Binaries available for Apple Silicon (arm64) and Intel (x64), plus an iOS XCFramework; KleidiAI support on Apple Silicon is disabled.
- Linux: Builds provided for Ubuntu x64, arm64, and s390x CPUs, as well as Vulkan, ROCm 7.2, OpenVINO, SYCL FP32, and SYCL FP16 variants.
- Android: An arm64 CPU binary is available.
- Windows: Binaries support x64 and arm64 CPUs, OpenCL Adreno, CUDA 12 and 13, Vulkan, OpenVINO, SYCL, and HIP.
- openEuler: Support for x86 (310p, 910b) and aarch64 (310p, 910b) with ACL Graph is included, while the pull request for this platform remains disabled.
This release provides updated binaries across multiple platforms and hardware accelerators, ensuring broader compatibility and stability for users running llama.cpp.