Ai model of Openai and Google Deepmind Get the gold medal score in the International Olympics 2025 (IMO), one of the oldest and most challenging school math competitions, the company independently in the present.
The result of the Activate only how fast AI system is, and yet, how to match Google and Opening appear in the AI races. AI competing companies for general perceptions to advance the AI race: Battle “vibes” can have great implications to achieve the biggest AI talent. Many AI researchers come from the background in competitive mathematics, as benchmarks like imo means more than others.
Last year, Google Print Silver Medals in Imo Using the “official” system, means human need to translate problems to machine readable format. This year, two opens and Google typing “informally to the competition, which can educate the question and return the answers based on the evidence. The two companies are expressing ai model that made higher than higher school students and Google Ai models since last year, without requires a human translation.
In interview with TechCrunch, researchers at the back of OPNAIA and Google’s Especially the gold medal performances are contributed around the mid-AI model in unrelocial domains. While the AI model that tends to do the correct answer, such as mathematical tasks or system coding, such as better cursing purchases, such as buying complex research.
However, Google enhances the questions about the opening of Opnaai and announced the performance of gold medal IMO performance. After all, if you will enter a AI model into a math contest for high school, you may show you like teenagers.
Shortly after the opening announced the birthday of Saturday morning, Google DeepMind’s CEO and researchers take social media to Apenai Slam to announce the Medan Mervous Premature – Shortly after I declare the high school has won competition on Friday night – and do not have an official model test evaluated by IMO.
Thang Luong, Google Deepmind Senior senior researchers, the Techcrunch tech was aware of Google to honor Imo caused the students involved in the competition.
TechCrunch Events
San Francisco
|
October 27, 2025
Luong says Google has been working on the mera organizer since the last year to prepare the test and want to have a blessing and a formal grading Announced the official results, which is done on MondaySee rankings-.
“IMO organization has a grading guidelines,” Luong said. “Being an evaluation of any not based on the guidelines cannot claim about gold-medali levels (performance).”
Noam Brown, Opening Opening Opening Researchers are working in the IMO model, informing the techo to opener several months, but it can be used in natural language system, due to working with a natural language system that is expected to be more useful. Brown says Opening doesn’t know about me to do informal tests with Google.
Opening says Third Party Rent – Three Smooth Imo knows the grading system – for the performance of the AI model. After the opening of the study of gold medal, Brown said that the company approached you, who then told the company to wait until the Award’s Award’s ceremony.
Imo does not respond to techcrro requests for comments.
Google should not be wrong here – it’s more right, more rigorous to reach the gold medal score – but the debate may miss a larger image: A few models of a few AI aircraft. The state of the world is delivered the brightest student to compete in IMO this year, and only a few percent you do as well as the opening and Google model model.
When the opening used to have an important lead for the industry, it should be felt that the race is more suitable than the company you want to be admitted. Opening is expected to release GPT-5 in the coming months, and the company is always hopeless to avoid impression if it still causes industrial AI.