Gemini vs. Grock vs. Cloud vs. Deepsek: Which tool is the best for chess? , Chess news

The opening day of the AI chess exhibition tournament hosted by Google’s Kaggle game Erina Project was seen by four big language models (LLM), who was winning a major 4–0 win to proceed in the semi-finals. Gemini 2.5 Pro, O 4-Mune, Grock 4, and O3 defeated their respective opponents Cloud 4 Opus, Dipsek R1, Gemini 2.5 flash, and km 2, showing the abilities of general-affected AI models in the strategic gameplay.A new initiative of Google -owned Kaggle, the purpose of Kaggle Game Arena is to evaluate how LLM performs in a competitive environment. Eight major LLMs in the tournament are competing in a single-ruling knockout bracket, with the games on several platforms.Go beyond the border with our YouTube channel. Subscribe now!Google has partnered with lamps to organize this unique tournament, where LLM uses a universal controller called “Harnes” to imagine and make tricks. Each AI has four attempts to maintain a legal step, which fails resulting in losing the game.The match between Km K2 and O3 ended quickly, out of which no game was beyond eight moves. Despite showing the ability to follow the opening theory for initial tricks, the km K2 continuously failed to perform legal tricks.The O4-Mune victory against Deepsek R1 displayed a pattern of strong opening moves after a decline in sports quality. Despite discrepancies, the O4-Min managed to receive two checkmates during the match.Responding to the impressive performance of Groke 4 in the tournament, Elon Musk posted on X, “This is a side effect BTW. @Xai did not make almost any attempt on chess.”The Mithun 2.5 Pro match against Cloud 4 Opus was more checkmate than illegal moves. The first game showed both AIs to nine until nine, when Claude 4 Opus made a significant error with 10 … G5.Groke 4 gave the strongest performance of the day, which shows special skills in the identification and capitalization of the unimpressed pieces in its match against Gemini 2.5 flash.The tournament has revealed three primary challenges for LLM in chess: imagining the entire board, understanding pieces, and doing legal tricks. These boundaries vary between different AI models.The competition continues on Wednesday, August 6, which starts at 1 pm / 19:00 cest / 10:30 PM IST. Viewers can watch this incident live on twitch and YouTube channels of GM Hikaru Nakamura, as well as on the dedicated event page of the tournament.