%20--%3e%3cdefs%3e%3cstyle%3e%20.st0%20{%20fill:%20%23061b40;%20}%20.st1%20{%20fill:%20%23306af1;%20}%20.st2%20{%20fill:%20%235ce5cf;%20}%20%3c/style%3e%3c/defs%3e%3cg%3e%3cpath%20class='st0'%20d='M55,10.5h9v3h-9v9h12V7.5h-12v3ZM64,19.5h-6v-3h6v3Z'/%3e%3cpolygon%20class='st0'%20points='69%2016.5%2078%2016.5%2078%2019.5%2069%2019.5%2069%2022.5%2081%2022.5%2081%2013.5%2072%2013.5%2072%2010.5%2081%2010.5%2081%207.5%2069%207.5%2069%2016.5'/%3e%3cpolygon%20class='st0'%20points='95%2010.5%2095%207.5%2083%207.5%2083%2022.5%2095%2022.5%2095%2019.5%2086%2019.5%2086%2016.5%2095%2016.5%2095%2013.5%2086%2013.5%2086%2010.5%2095%2010.5'/%3e%3cpath%20class='st0'%20d='M40,1.5v21h11.6l1.4-1.4v-7.6h0c0,0-1.4-1.5-1.4-1.5l1.4-1.4V2.9l-1.4-1.4h-11.6ZM50,19.5h-7v-6h7v6ZM50,10.5h-7v-6h7v6Z'/%3e%3c/g%3e%3cpath%20class='st1'%20d='M23.1,24L14.7,4.8l-4.9,11.2h3.8l-1.8,4H3.3L12.1,0H2C.9,0,0,.9,0,2v20c0,1.1.9,2,2,2h21.1Z'/%3e%3cpath%20class='st2'%20d='M34,0h-16.8l10.6,24h6.2c1.1,0,2-.9,2-2V2C36,.9,35.1,0,34,0ZM32.5,20h-4V4h4v16Z'/%3e%3c/svg%3e)
In the wave of global AI research, Google's first large model competition has attracted widespread attention. The competition will be held from August 5 to 7 at Kaggle Game Arena, bringing together eight top AI models, including DeepSeek and Kimi, competing on the stage of chess in an intense battle.
The participating models include o4-mini from OpenAI, DeepSeek-R1, Kimi K2Instruct, Gemini2.5Pro (Google), Claude Opus4 (Anthropic), Grok4 (xAI), and Gemini2.5Flash. Each model represents the cutting-edge technology in the current AI field. The organizers have specially invited world-class chess experts to provide commentary, adding professionalism and entertainment to the competition.
The organizers of the competition stated that the initial purpose of this competition is to evaluate the performance of AI models in real competitive environments. With the rapid development of AI technology, existing benchmark testing methods are no longer effective in distinguishing the true capabilities of models. Kaggle Game Arena was established to solve this problem. By engaging in actual confrontations in strategy games, researchers can more comprehensively evaluate the performance of models.
The competition will use a round-robin format to ensure the reliability of the statistical results. Each pair of models will compete in multiple matches, and the final ranking will be strictly evaluated based on the match results. To ensure transparency, the execution framework and environment of the competition will be fully open source, allowing the audience to view the match schedule and progress in real time.
The competition will adopt a single-elimination format. Each match consists of four games, and the model that first scores two points will advance. If the game ends in a draw, both models will play a tiebreaker game. During the competition, each model will face challenges with text input and cannot use external tools such as chess engines for assistance, increasing the complexity and interest of the competition.
Demis Hassabis, co-founder of Google DeepMind, said: "Games have always been an important test ground for AI capabilities. We are extremely excited about Kaggle Game Arena's potential to drive AI advancement. As more games and challenges are added, AI capabilities are sure to improve rapidly!"
As the competition approaches, audiences are full of anticipation for the final results, passionately discussing which model will stand out in this competition. Regardless of the outcome, this competition will bring new ideas to the evaluation methods of AI models and promote continuous technological advancement.
