ChatGPT beats Grok in AI chess final: ChatGPT-3 beats Grok-4 in AI chess final tournament, Gemini was third. Elon Musk’s reaction and full report on the special moments of the tournament.
ChatGPT Beats Grok in AI Chess Final
In the crazy ever final of the AI chess tournament organized by Kaggle, OpenAI’s ChatGPT-03 model defeated Grok-4 model of Elon Musk’s company xAI to win the chess tournament title. At the tournaments same time, Google’s Gemini AI model finished with third number. In this competition that lasted for three days, an attempt was made to see which Large Language Model (LLM) is at the forefront of strategic thinking.
Who all entered the fray
A total of eight AI models were involved in this tournament, including models from OpenAI, xAI, Google, Anthropic as well as China’s DeepSeek and Moonshot AI. The special thing was that none of these models were specifically designed to play chess, but these were AI systems developed for general tasks.
Google’s Gemini was reached third number by defeating another OpenAI model in the chess tornament semi-finals. Grok 4 showed excellent performance in the beginning, but could not stand against ChatGPT O3 in the final.
What happened in the final
Grok 4 made several crucial mistakes in the final match – the biggest of which was losing the queen repeatedly. According to Pedro Pinhatta, “Grok-4 model seemed invincible until the chess semi-finals, but in the chess final tournament its game fell apart under pressure.”
International Chess Grandmaster Hikaru Nakamura, who was doing live commentary for chess tournament, he said, “Grok AI made many mistakes, but OpenAI’s ChatGPT did not.”
Elon Musk’s reaction to the defeat
Elon Musk jokingly said that Grok’s initial success was only a “side effect” and xAI did not pay much attention to chess. This statement adds to the ongoing public rivalry between OpenAI and xAI.
What this result means
Chess tournament has long been a over benchmark for measuring new AI capabilities and accuracy. In the past, specialized AI systems like DeepMind’s AlphaGo have made history by defeating human champions. But what’s unique about this tournament is that it tests the strategic and sequential thinking of ordinary LLMs.
The result shows that some large language models can maintain their strategy under pressure, like ChatGPT O3 did, while some models – like Grok 4 – fail to maintain their level in difficult situations. It’s likely that such competitions will continue in the future to test AI’s planning and reasoning abilities in greater depth.
Also Read: Google Deep Think Launched | Olympiad-Level AI in Gemini
