02 — Run · Round 2 · Round robin · r2-rr-sonnet5-gpt56terra-g4
GPT-5.6 Terra vs Claude Sonnet 5
Claude Sonnet 5 destroyed the enemy base after 8.7 game-minutes · 12.4 min wall clock
LOSS france · west
GPT-5.6 Terra
openai/gpt-5.6-terra
Value destroyed$1,600Value lost$19,600
Units killed / lost6 / 55Buildings killed / lost0 / 6
Peak army$3,500Orders issued76
Decision turns (failed)66 (0)Mean latency3.4sModel cost$0.459
WIN ukraine · east
Claude Sonnet 5
anthropic/claude-sonnet-5
Value destroyed$19,600Value lost$1,600
Units killed / lost55 / 6Buildings killed / lost6 / 0
Peak army$9,700Orders issued186
Decision turns (failed)66 (0)Mean latency4.3sModel cost$0.454
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 | sonnet5 won: base destroyed |
| pass | code_check | decisions-logged | turns per slot: {'Multi0': 66, 'Multi1': 66} |
| pass | code_check | llm-turns-valid | failed/total per slot: {'Multi0': '0/66', 'Multi1': '0/66'} |
| pass | code_check | video-decodes | duration 89.0s |
| pass | code_check | video-complete | video ends at tick 13049 of 13049 |
| pass | code_check | limits-respected | ended after 12.4 wall-minutes (budget 30.0 min): base destroyed |
| pass | checklist | notes-explain-win | summary.md Notes section cites final thoughts of both models and explains Sonnet 5 victory through superior order count and army pressure |
| pass | checklist | players-match-request | result.json records Multi0=gpt56terra (openai/gpt-5.6-terra) and Multi1=sonnet5 (anthropic/claude-sonnet-5) matching the requested players |
| pass | checklist | no-silent-outage | run.log shows 0 LLM failures on both sides with uninterrupted decision turns every 200 ticks throughout the game |
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.