The llama.cpp b9842 release introduces a change to deduplicate preset and cached model entries in the /v1/models endpoint. This update is signed off by Adrien Gallouët from Hugging Face.
- macOS Apple Silicon (arm64) binaries are available, while KleidiAI support remains disabled.
- Linux builds include Ubuntu x64/arm64/s390x CPU versions, Vulkan, ROCm 7.2, OpenVINO, and SYCL FP32/FP16 variants.
- Android arm64 (CPU) binaries are provided for mobile deployment.
- Windows releases cover x64/arm64 CPU, OpenCL Adreno, CUDA 12.4/13.3, Vulkan, OpenVINO, SYCL, and HIP backends.
- openEuler support includes x86 and aarch64 builds with ACL Graph for 310p and 910b chips, though standard openEuler is disabled.
- A standalone UI binary is also included in the release assets.
This release provides updated binaries across multiple platforms and hardware accelerators, ensuring compatibility with various CPU and GPU architectures.