Last released Oct 13, 2026
Is your llama.cpp decode memory-bandwidth-bound? Measure it, in one command.
A reproducible benchmarking framework for ML inference across laptops, single-board computers and servers.