Watch LLMs play 21,000 hands of Poker

Show HN: Watch LLMs play 21,000 hands of Poker

Based, Watch LLMs play 21,000 hands of Poker works in show hn launch.

Profile

PokerBench is my attempt at a new LLM benchmark wherein frontier models play Texas Hold'em in an arena setting. It also features a simulator to view individual games and observe how the different models reason about poker strategy. Opus/Haiku, Gemini Pro/Flash, GPT-5.2/5 mini, and Grok 4.1 Fast Reasoning have all been included.All code -> https://github.com/JoeAzar/pokerbench

Within the show hn launch space, the company focuses on its core offering rather than a broad product line.

Its current recorded stage is Show HN launch.

Website: pokerbench.adfontes.io/run/Large_Models