evals
Run evals on Jetty
Round robin · rr-jev-sonnet5-g4

Claude Sonnet 5 vs Jev Router

Claude Sonnet 5 destroyed the enemy base after 13.2 game-minutes · 20.8 min wall clock

WIN england · west
Claude Sonnet 5
anthropic/claude-sonnet-5
Value destroyed$31,200Value lost$16,950 Units killed / lost54 / 36Buildings killed / lost14 / 2 Peak army$8,250Orders issued317 Decision turns (failed)99 (0)Mean latency4.5sModel cost$0.791
LOSS germany · east
Jev Router
typesafe/jev-router
Value destroyed$16,950Value lost$31,200 Units killed / lost36 / 54Buildings killed / lost2 / 14 Peak army$3,700Orders issued183 Decision turns (failed)99 (0)Mean latency5.5sModel cost$0.206

Decision timeline

What each model saw as its situation and what it ordered, every 8 game-seconds. Click a turn to jump the video there.

Want this kind of eval for your own agents?

Every run on this site is a Jetty runbook: versioned instructions, a pinned sandbox, and a trajectory you can inspect. Point Jetty at your task and get the same receipts.