Last released May 13, 2026
Local-first CLI + web dashboard for benchmarking LLMs across quality, speed, reliability, and a real multi-turn agent loop. Hardware-aware, deterministic, reproducible.
Supported by
