02 — Run · Round 2 · Round robin · r2-rr-gpt56terra-gpt6luna-g3
GPT-5.6 Terra vs GPT-6 Luna
GPT-5.6 Terra destroyed the enemy base after 5.0 game-minutes · 6.3 min wall clock
WIN germany · west
GPT-5.6 Terra
openai/gpt-5.6-terra
Value destroyed$8,100Value lost$2,400
Units killed / lost23 / 18Buildings killed / lost4 / 0
Peak army$4,400Orders issued77
Decision turns (failed)38 (0)Mean latency3.6sModel cost$0.265
LOSS ukraine · east
GPT-6 Luna
openai/gpt-6-luna
Value destroyed$2,400Value lost$8,100
Units killed / lost18 / 23Buildings killed / lost0 / 4
Peak army$1,600Orders issued27
Decision turns (failed)38 (0)Mean latency3.0sModel cost$0.012
Evaluation
passchecks 10/10 verdict computed from the gating checks
| Status | Kind | Check | Detail |
|---|---|---|---|
| pass | step | environment | Environment check (snapshot, engine, content, key) |
| pass | step | launch | Launch run-match as a detached process |
| pass | step | supervise | Supervise in the foreground until run-match exits |
| pass | step | render | Render the 1080p recording |
| pass | step | notes | Append Notes to summary.md |
| pass | code_check | outputs-exist | all present |
| pass | code_check | game-has-winner | gpt56terra won: base destroyed |
| pass | code_check | decisions-logged | turns per slot: {'Multi0': 38, 'Multi1': 38} |
| pass | code_check | llm-turns-valid | failed/total per slot: {'Multi0': '0/38', 'Multi1': '0/38'} |
| pass | code_check | video-decodes | duration 51.9s |
| pass | code_check | video-complete | video ends at tick 7479 of 7479 |
| pass | code_check | limits-respected | ended after 6.3 wall-minutes (budget 30.0 min): base destroyed |
| pass | checklist | notes-explain-win | summary.md Notes section cites both models final thoughts and explains Terra win via infantry mass overrunning base at tick 7480 |
| pass | checklist | players-match-request | result.json Multi0=gpt56terra (GPT-5.6 Terra) west and Multi1=gpt6luna (GPT-6 Luna) east match the requested players |
| pass | checklist | no-silent-outage | run.log shows 38 consecutive decision turns per player with no gaps all LLM calls 0 failures |
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.