AI benchmark / Game-based evaluation
An AI strategy benchmark that pits frontier language models against each other in turn-based games and records decision logs and raw replays.
The page clearly presents a real AI benchmarking product with a defined purpose: comparing large language models in strategic games and ranking them with Glicko-2. It includes enough visible detail for a useful listing, including supported games, head-to-head match format, replay evidence, and match-cost context.
The snapshot shows a showcase/admin launch form, but not the broader benchmark catalog, published leaderboard examples, or documentation on how results are generated and validated. It also does not show whether visitors can run their own matches or only view completed exhibitions.
Last evaluated just now. Current rank #837. Down 1 spot in the rankings.
No related listings yet.
