OpenAI's o3 faced off against xAI's Grok 4 in a chess tournament. It swept the board.
Getty Images
- OpenAI's o3 swept xAI's Grok 4 in the final of an AI chess tournament.
- Hosted by Google's Kaggle, the tournament pitted general-purpose AI models against each other.
- The win adds to the ongoing public rivalry between Sam Altman and Elon Musk.
It might as well have been the tech industry's latest grudge match.
On one side: OpenAI's reasoning-tuned o3 model. On the other, Elon Musk's flagship AI model, Grok 4.
In an AI chess tournament hosted by Google's Kaggle, the two met in a final that put their skills — and their makers' reputations — to the test.
o3 swept the board 4—0 and took the AI chess crown.
The third-place playoff went to Google's Gemini 2.5 Pro, which defeated o4-mini 3.5—0.5
Tech companies have long used chess to measure a computer's progress and capabilities, and today's specialist engines are virtually unbeatable, even for the best human players.
In 2016, AlphaGo — a program developed by Google's DeepMind to play the Chinese strategy game Go — made headlines when it defeated world champion Lee Sedol.
But this competition pitted general-purpose AI programs against each other to see which model would come out on top over the chessboard.
OpenAI, xAI, and Kaggle did not respond to a request for comment from Business Insider.
How the showdown unfolded
The tournament ran from August 5 to August 7, in a knockout format featuring eight large language models: OpenAI's o3 and o4-mini, xAI's Grok 4, Google's Gemini 2.5 Pro and Gemini 2.5 Flash, Anthropic's Claude 4 Opus, DeepSeek's DeepSeek R1, and Kimi's k2.
Grok 4 impressed early, cruising through its bracket to reach the final. o3 did the same, dismantling every opponent along the way.
When the two met in the last match, it wasn't close.
"Up until the Semifinals, it seemed like nothing would be able to stop Grok 4 on its way to winning the event," wrote Pedro Pinhata, a writer for Chess.com, which provided coverage for the tournament.
"Despite a few moments of weakness, X's AI seemed to be by far the strongest chess player," he added.
But once the final began, that prediction fell apart. "The chatty o3 simply dismantled its mysterious opponent with four convincing wins," Pinhata wrote.
"Grok's play was unrecognizable, blundering soon and often. And for the most part, o3 showed no mercy."
Former world champion Magnus Carlsen didn't hold back during a livestream of the final on YouTube Channel "Take Take Take."
He likened Grok to "that one guy in a club who has learnt theory and literally knows nothing else."
"It just makes the worst blunders," he said.
Altman vs Musk
OpenAI's win over Grok is another chapter in an increasingly public feud between its cofounders.
Musk on Monday threatened to sue Apple, alleging App Store favoritism toward OpenAI. Altman shot back by pointing to Musk's control of X.
Musk posted a ChatGPT 5 Pro exchange on Tuesday in which the bot deemed him "more trustworthy" than Altman.
Back in May, Altman tried the same trick with Grok, asking who should lead the AI race if humanity's fate was at stake.
"If forced, I'd lean toward Musk for his safety emphasis, critical for humanity's survival, though Altman's accessibility is vital," Grok said in response to Altman's query. "Ideally, their strengths should combine with regulation to ensure AI benefits all."