Last released Jul 18, 2026
Auto-tune and benchmark llama.cpp / ik_llama.cpp inference on NVIDIA, AMD (ROCm), and Apple Silicon GPUs
Supported by