OpenAI’s ChatGPT-powered o3 crushed Elon Musk’s xAI Grok 4 in a chess AI showdown. The tournament, held on Google-owned Kaggle, pitched everyday use AI models—not dedicated chess engines—against each other.
OpenAI’s o3 went undefeated and beat Grok 4 in the final match, ending xAI’s surprising run. Google’s Gemini snagged third place after topping another OpenAI model.
Grok made critical blunders in the final, like repeatedly losing its queen. Pedro Pinhata from Chess.com said: 
“Up until the semi finals, it seemed like nothing would be able to stop Grok 4 on its way to winning the event.”
“Despite a few moments of weakness, X’s AI seemed to be by far the strongest chess player… But the illusion fell through on the last day of the tournament.”
“Grok made so many mistakes in these games, but OpenAI did not.”
Chess grandmaster Hikaru Nakamura echoed the take during his final day livestream:
“Grok made so many mistakes in these games, but OpenAI did not.”
Before the final, Musk posted on X that xAI’s chess success was just a “side effect” since it “spent almost no effort on chess.”
Eight large language models from top developers including Anthropic, Google, DeepSeek, and Moonshot AI battled across three days to test reasoning and strategy skills using chess—a classic AI benchmark.
While AI chess machines once focused solely on chess, this event spotlighted the evolving capabilities of general AI models tackling complex rule-based tasks on the fly.
The tournament adds a fresh chapter to the ongoing rivalry between OpenAI and Musk’s xAI, both claiming their latest models are the smartest around.
See full coverage on Chess.com and Musk’s X post.
