COGNITIVE_ARENA: EDGE_BENCHMARK

Run identical prompts side-by-side across two local models. Compare real-time token throughput (t/s), latency, and accuracy offline.

Engine A (Left Console)STANDBY

Select Architecture

Engine B (Right Console)STANDBY

Select Architecture