Last released Nov 10, 2026
One serving daemon for local LLM inference, across runtimes and machines
Fast local LLM inference on Apple Silicon and NVIDIA GB10: the basert CLI, engine, and Python bindings