10 Sources
10 Sources
[1]
Gemini AI solves coding problem that stumped 139 human teams at ICPC World Finals
Like the rest of its Big Tech cadre, Google has spent lavishly on developing generative AI models. Google's AI can clean up your text messages and summarize the web, but the company is constantly looking to prove that its generative AI has true intelligence. The International Collegiate Programming Contest (ICPC) helps make the point. Google says Gemini 2.5 participated in the 2025 ICPC World Finals, turning in a gold medal performance. According to Google this marks "a significant step on our path toward artificial general intelligence." Every year, thousands of college-level coders participate in the ICPC event, facing a dozen deviously complex coding and algorithmic puzzles over five grueling hours. This is the largest and longest-running competition of its type. To compete in the ICPC, Google connected Gemini 2.5 Deep Think to a remote online environment approved by the ICPC. The human competitors were given a head start of 10 minutes before Gemini began "thinking." According to Google, it did not create a freshly trained model for the ICPC like it did for the similar International Mathematical Olympiad (IMO) earlier this year. The Gemini 2.5 AI that participated in the ICPC is the same general model that we see in other Gemini applications. However, it was "enhanced" to churn through thinking tokens for the five-hour duration of the competition in search of solutions. At the end of the time limit, Gemini managed to get correct answers for 10 of the 12 problems, which earned it a gold medal. Only four of 139 human teams managed the same feat. "The ICPC has always been about setting the highest standards in problem-solving," said ICPC director Bill Poucher. "Gemini successfully joining this arena, and achieving gold-level results, marks a key moment in defining the AI tools and academic standards needed for the next generation." More than human At the ICPC, only correct solutions earn points, and the time it takes to come up with the solution affects the final score. Gemini reached the upper rankings quickly, completing eight problems correctly in just 45 minutes. After 677 minutes, Gemini 2.5 Deep Think had 10 correct answers, securing a second-place finish among the university teams. You can take a look at all of Gemini's solutions on GitHub, but Google points to Problem C as especially impressive. This question, a multi-dimensional optimization problem revolving around fictitious "flubber" storage and drainage rates, stumped every human team. But not Gemini. According to Google, there are an infinite number of possible configurations for the flubber reservoirs, making it challenging to find the optimal setup. Gemini tackled the problem by assuming that each reservoir had a priority value, which allowed the model to find the most efficient configuration using a dynamic programming algorithm. After 30 minutes of churning on this problem, Deep Think used nested ternary search to pin down the correct values. Gemini's solutions for this year's ICPC were scored by the event coordinators, but Google also turned Gemini 2.5 loose on previous ICPC problems. The company reports that its internal analysis showed Gemini also reached gold medal status for the 2023 and 2024 question sets. Google believes Gemini's ability to perform well in these kinds of advanced academic competitions portends AI's future in industries like semiconductor engineering and biotechnology. The ability to tackle a complex problem with multi-step logic could make AI models like Gemini 2.5 invaluable to the people working in those fields. The company points out that if you combine the intelligence of the top-ranking university teams and Gemini, you get correct answers to all 12 ICPC problems. Of course, five hours of screaming-fast inference processing doesn't come cheap. Google isn't saying how much power it took for an AI model to compete in the ICPC, but we can safely assume it was a lot. Even simpler consumer-facing models are too expensive to turn a profit right now, but AI that can solve previously unsolvable problems could justify the technology's high cost.
[2]
Gemini just aced the world's most elite coding competition - what it means for AGI
A Gemini model won gold at a challenging coding competition.The model correctly answered 10 out of 12 problems.The win could have major implications for AGI, says Google. In recent years, large language models (LLMs) have become an integral part of many software developers' toolkits, helping them build, refine, and deploy apps more quickly and effectively. Now, Google says that one of its most advanced models has achieved a major coding breakthrough that could help lead to new scientific discoveries -- including, potentially, the attainment of artificial general intelligence, or AGI. Also: Will AI think like humans? We're not even close - and we're asking the wrong question Gemini 2.5 Deep Think, a state-of-the-art version of Google's flagship AI model that uses advanced reasoning capabilities to break problems down into multiple components, has achieved gold medal performance at the 2025 International Collegiate Programming Contest (ICPC) World Finals, the company announced Wednesday. Google wrote in a blog post that the "advanced version" of Gemini 2.5 Deep Think operates as a kind of automated and integrated team. "To tackle a problem, multiple Gemini agents each propose their own solutions using terminals to execute code and tests, and then iterate the solutions based on all of the attempts," the company wrote. The ICPC is widely recognized as the world's most prestigious and difficult university-level coding competition. Teams hailing from close to 3,000 universities across 103 countries competed in this year's finals, which were held Sept. 4 in Baku, Azerbaijan. Each team must solve a set of complex problems within a five-hour time period. There's no room for error: Only perfect answers get points. Also: I did 24 days of coding in 12 hours with a $20 AI tool - but there's one big pitfall Gemini correctly solved 10 out of the 12 problems in this year's ICPC finals, achieving a gold medal-level performance and the second-highest score overall compared to a group of human contestants. Gemini 2.5 Deep Think, along with an experimental reasoning model from OpenAI, also achieved gold medal-level performance at this year's International Mathematical Olympiad, the companies announced in July. "Together, these breakthroughs in competitive programming and mathematical reasoning demonstrate Gemini's profound leap in abstract problem-solving -- marking a significant step on our path toward artificial general intelligence (AGI)," Google wrote in its blog post. In what Google describes in a blog post as "an unprecedented moment," Gemini quickly and correctly solved one of the 12 problems in the competition that stymied all of the human competitors. There were two problems that it didn't manage to solve, on the other hand, which other teams did successfully. Also: OpenAI has new agentic coding partner for you now: GPT-5-Codex The third problem in the challenge, Problem C, asked competitors to devise a solution for distributing liquid through a series of interconnected ducts, so that reservoirs connected to each duct would be filled as quickly as possible. Each duct could be closed, open, or partially open, meaning there was an infinite number of possible configurations. In its search for the optimal configuration, Gemini took a surprising approach: It began by assigning a numerical value to each reservoir to determine the priority it should be assigned relative to the others. The model then deployed an algorithm and a game-theoretical concept known as the minimax theorem to find a solution. The whole process took less than half an hour. No human competitor was able to solve it. Also: I built a business plan with ChatGPT and it turned into a cautionary tale Although less monumental in its significance, this kind of problem-solving capability is reminiscent of the famous Move 37 during AlphaGo's 2016 game against Go world champion Lee Sedol, in which that AI model (developed by Google DeepMind) adopted a strategy that surprised human experts in the moment, but turned out to be decisive to its victory. Since then, "Move 37" has become shorthand for moments in which AI acts in creative or unexpected ways which challenge our conventional norms of intelligent problem-solving. Gemini's top-tier performance at the 2025 ICPC has implications far beyond software development, according to Google. "The skills needed for the ICPC -- understanding a complex problem, devising a multi-step logical plan, and implementing it flawlessly -- are the same skills needed in many scientific and engineering fields, such as designing new drugs, or microchips," the company wrote in its blog post, saying that this development shows AI could help solve difficult problems for the benefit of humanity (a familiar pseudo-promise AI companies often make). Also: Google's new open protocol secures AI agent transactions - and 60 companies already support it The notion that AI could eventually assist with scientific discovery has long been a dream for many computer scientists. Earlier this month, OpenAI launched an internal initiative aimed at this very goal. Earlier this month, Harvard Medical School designed an AI model that could help target degenerative disease and cancer treatment. According to Google, the best path forward in this regard will likely be some form of human-AI collaboration, through which advanced agentic models like Gemini 2.5 Deep Think suggest novel solutions to particularly difficult technical problems.
[3]
DeepMind achieves gold at 'coding Olympics' in AI milestone
Google DeepMind's latest artificial intelligence model has achieved a "gold-medal level" performance at a competition known as the "coding Olympics", in what the group describes as a milestone in the development of the revolutionary technology. The London-based laboratory run by British Nobel laureate Sir Demis Hassabis said on Wednesday that its AI reasoning model, Gemini 2.5 Deep Think, achieved the result against the best human competitors at the International Collegiate Programming Contest (ICPC) World Finals in early September. The competition is considered the most prestigious programming contest in the world. Former participants include Google co-founder Sergey Brin and OpenAI's chief scientist Jakub Pachocki. DeepMind said the Gemini model's performance would have ranked second overall in the competition. It was also able to solve a problem that no human competitor could solve. The breakthrough comes as the newest generation of AI models are increasingly being used by software engineers to assist with computer programming. Meanwhile, DeepMind's technology has already been used to win against humans in other elite competitions, from beating the world's best player at the board game Go, to achieving gold at the International Mathematical Olympiad. Quoc Le, vice-president and Google fellow said: "This is a historic moment towards AGI," referring to artificial general intelligence -- systems that surpass human capabilities -- a major goal for AI researchers for decades. "It's impressive for a purely AI system with no human in the loop to be able to get the performance that they did," said Jelani Nelson, the chair of University of California, Berkeley's electrical engineering and computer sciences department, who has coached several ICPC teams at Massachusetts Institute of Technology, Harvard and UC Berkeley. "If someone had told me just a few years ago that we would have new technology that was able to perform at this level in math and in computer science, I would not have believed them," added Nelson. In the coding competition, teams of three are given one computer with which to solve 12 hard programming problems in five hours. Teams are ranked on speed, accuracy and the number of questions they answer. This year, competitors were able to solve 10 out of the 12 questions. From the 139 competing teams this year, only four teams won gold medals. To solve the problems, participants have to understand complex problems, have a logical plan to solve them, and execute them without errors. Hard maths problems also require abstract reasoning skills and creativity. DeepMind's AI tool had a crucial advantage over people: it did not have to work in a team. "When I coach my teams, the assumption is that I don't have to teach them how to solve problems . . . I can only give them advice on how to work together in a stressful situation," said Bartek Klin, an associate professor of computer science of the University of Oxford, and an ICPC coach. The DeepMind team used "reinforcement learning" -- a technique that rewards AI systems for desired outcomes -- to train its Gemini model further with very hard maths, reasoning and coding problems. Competitive coding is the "ultimate thinking game", because it requires models to come up with new approaches and generalise learnings, instead of just memorising solutions, said Heng-Tze Cheng, research director and principal scientist at Google DeepMind. But Oxford university's Klin said success in a competitive coding environment that prioritises speed does not necessarily translate to great software development in practice. "In real life, the hardest problems are the ones that take half a year to think about," he said. While the Gemini model was able to solve a problem the competitors were not, it was also not able to solve all the tasks that some of its human counterparts did complete. DeepMind said the experiment showed how AI models could "provide unique, novel contributions that complement the skills and knowledge of human experts". Le says the advancement also has the potential to transform many scientific and engineering disciplines, which require mathematical understanding and coding, such as designing new drugs and computer chips. "Solving math and computer competitive coding is a key step to understanding how our intelligence works," said Le.
[4]
OpenAI and DeepMind AI outperform top students in global coding contest
Serving tech enthusiasts for over 25 years. TechSpot means tech analysis and advice you can trust. What just happened? OpenAI and DeepMind have tested their systems in many high-level contests. DeepMind's AI has defeated world champions in Go and secured gold medals in the International Mathematical Olympiad, while OpenAI reported a win in mathematics at the same Olympiad this summer. The results, together with their recent achievements in coding's most recognized competition, show how these settings are becoming proving grounds for technologies approaching the frontier of human-level reasoning. Artificial intelligence from Google DeepMind and OpenAI has reached a new benchmark in competitive programming, with both groups reporting that their latest models would have placed at the top of this year's International Collegiate Programming Contest World Finals. While neither company officially entered the event, held in early September, internal tests suggest that OpenAI's GPT-5 model would have finished first, while DeepMind's newly trained Gemini 2.5 Deep Think system would have ranked second. The ICPC has produced some of the most influential figures in the technology industry, including Google co-founder Sergey Brin and OpenAI Chief Scientist Jakub Pachocki. Teams of three students, working at a single computer within a five-hour window, must solve twelve programming problems that require abstract reasoning, creative problem-solving, and error-free execution. This year, the strongest human competitors managed to solve ten questions; OpenAI reported that GPT-5 completed all twelve, with eleven correct on its first attempt. DeepMind's Gemini 2.5 also matched and outperformed many human participants, solving one task that no team of students could complete. The achievement underscores how closely artificial intelligence systems are now competing with elite human programmers in areas once thought beyond reach. "This is a historic moment towards AGI," Quoc Le, vice president of Google DeepMind and Google Fellow, told The Financial Times. Demis Hassabis For OpenAI, the result highlights the increasing sophistication of its GPT-5 system, which the company used on all problems except the final, most complex one. That last problem was solved using GPT-5 in tandem with an experimental reasoning model still under development. London-based DeepMind, founded by British neuroscientist and chess prodigy Sir Demis Hassabis, selected a different path. It combined reinforcement learning - an approach that rewards systems for producing correct results - with intensive exposure to difficult mathematics, reasoning exercises, and coding challenges to train Gemini 2.5 Deep Think. Experts in the programming contest community were struck by the demonstration. "It's impressive for a purely AI system with no human in the loop to be able to get the performance that they did," Jelani Nelson, chair of electrical engineering and computer science at the University of California, Berkeley, said. He noted that such capabilities seemed impossible until recently. "If someone had told me just a few years ago that we would have new technology able to perform at this level in math and computer science, I would not have believed them," Nelson, who has coached ICPC teams at Berkeley, Harvard, and MIT, added. Still, observers cautioned against drawing broad conclusions about AI's ability to write production-ready software. Bartek Klin, associate professor of computer science at the University of Oxford and an ICPC coach, said the competition rewards speed in high-pressure situations, a skill that does not necessarily translate into practical engineering success. "In real life, the hardest problems are the ones that take half a year to think about," Klin said. He noted that human teams must also develop strategies for collaboration, a challenge that AI systems do not face. DeepMind emphasized that Gemini 2.5's performance had not perfectly matched that of the leading human teams; in some cases, the AI did not solve problems that competitors completed successfully. Nonetheless, the laboratory highlighted the model's ability to produce unique solutions that no human team attempted. The company argued that this points to a future in which AI systems augment human intelligence by contributing original approaches to intractable problems. The companies see potential beyond contests. Le of DeepMind said that progress in mathematical reasoning and competitive coding could translate into breakthroughs in science and engineering. Disciplines such as drug design and semiconductor development, which require both algorithmic rigor and mathematical innovation, could ultimately benefit from models with demonstrated capacity to solve complex, abstract challenges. Heng-Tze Cheng, research director and principal scientist at DeepMind, described competitive programming as "the ultimate thinking game," since it requires developing new approaches rather than relying on memorized answers.
[5]
Google and OpenAI's coding wins at university competition show enterprise AI tools can take on unsolved algorithmic challenges
OpenAI and Google DeepMind demonstrated that their foundation models could outperform human coders -- and win -- showing that large language models (LLMs) can solve complex, previously unsolved algorithms. OpenAI's GPT-5 and Google's Gemini 2.5 Deep Think participated in the 2025 International Collegiate Programming Contest (ICPC) World Finals. The competition brings together coding teams from universities to compete in answering complex algorithmic questions. Although both models technically didn't compete alongside human teams -- their participation was governed by ICPC rules and supervised by the organizations -- the LLMs successfully answered problems that some contestants could not. GPT-5 managed to achieve a perfect score, answering 12 out of 12 problems, a performance akin to winning a gold medal in the event. Gemini 2.5 Deep Think solved 10 of the 12 algorithmic problems in 677 minutes, which Google DeepMind said in a blog post would rank second overall in the competition. OpenAI noted that they did not train a version of GPT-5 to learn how to answer questions at ICPC specifically. Google indicated that it entered an "advanced version" of Gemini 2.5 DeepThink. If you were wondering, the actual human gold medal winners of ICPC are teams from St. Petersburg State University, the University of Tokyo, Beijing Jiaotong University and Tsinghua University. (Harvard and MIT were the top-ranking American colleges, ending up on the silver medal level.) None of the human teams scored a 12 out of 12. The competition ICPC attracts thousands of participants, with 139 universities from at least 103 countries competing in the World Finals this year. During the finals, competitors must solve an identical set of algorithmic problems within a five-hour time frame. The final rankings will depend on which teams solved the questions and how quickly they were completed. "We officially competed in the onsite AI track of the ICPC, with the same 5-hour time limit to solve all twelve problems, submitting to the ICPC World Finals Local Judge - judged identically and concurrently to the ICPC World Championship submissions. We received the problems in the exact same PDF form, and the reasoning system selected which answers to submit with no bespoke test-time harness whatsoever. For 11 of the 12 problems, the system's first answer was correct. For the hardest problem, it succeeded on the 9th submission. Notably, the best human team achieved 11/12," OpenAI said in a post on X. Google, on the other hand, said Gemini "solved eight problems within just 45 minutes and two more problems within three hours." Additionally, Google said Gemini solved one problem that none of the university teams could figure out. It involved finding a way to distribute liquid through a series of ducts. "Gemini found an effective solution with a clever insight: it first assumed each reservoir has a 'priority value' representing how much each reservoir should be favored compared to the others. When given a set of priority values, the best configuration of the ducts can be found using a dynamic programming algorithm. Gemini discovered that by applying the minimax theorem, the original problem can be approached by finding the priority values that make the resulting flow most constrained. Leveraging the relationship between priority values and optimal flows, Gemini used nested ternary searches to quickly find optimal priority values in the bowl-like convex solution space, and solved Problem C," Google said. LLMs and complex problems There's no doubt that foundation models like GPT-5 and Gemini 2.5 can solve general knowledge questions; after all, these LLMs constantly prove their knowledge base on the more common benchmark tests available. What the performance at ICPC shows is that, given more complex math problems, and in a competitive coding event pitted against human coders, the models could beat humans. The gap has been narrowing for a while. Earlier this year, Google announced that Gemini won a gold medal at the International Mathematical Olympiad, one of the world's toughest math competitions. This performance comes just months after LLMs proved unable to answer complex math problems on the FrontierMath benchmark. Admittedly, some enterprise use cases do not need a model that can answer the world's most unsolvable programming questions. However, as enterprises find more and more complex workflows to delegate to AI systems and organizations seek more AI-powered analysis, having LLMs that have proven strong coding and mathematical skills would be extremely useful. It also shows that foundation models have come a long way, able to utilize deep abstract reasoning and creative problem-solving skills that may prove beneficial for enterprise issues in the future. A path to AGI Many believe that models displaying this level of reasoning and problem-solving represent a strong move towards artificial general intelligence. Closing the gap between human reasoning and LLMs via a programming competition certainly shows that the current crop of models is slowly marching down that path. These gold medal-winning performances have caught the attention of AI power users on social media.
[6]
Gemini achieves gold-level performance at the International Collegiate Programming Contest World Finals
Gemini 2.5 Deep Think achieves breakthrough performance at the world's most prestigious computer programming competition, demonstrating a profound leap in abstract problem solving. An advanced version of Gemini 2.5 Deep Think has achieved gold-medal level performance at the 2025 International Collegiate Programming Contest (ICPC) World Finals. This milestone builds directly on Gemini 2.5 Deep Think's gold-medal win at the International Mathematical Olympiad (IMO) just two months ago. Innovations from these efforts will continue to be integrated into future versions of Gemini Deep Think, expanding the frontier of advanced AI capabilities accessible to students and researchers. Solving complex tasks at these competitions requires deep abstract reasoning, creativity, the ability to synthesize novel solutions to problems never seen before and a genuine spark of ingenuity. Together, these breakthroughs in competitive programming and mathematical reasoning demonstrate Gemini's profound leap in abstract problem-solving -- marking a significant step on our path toward artificial general intelligence (AGI). The ICPC is globally recognized as the oldest, largest and most prestigious algorithmic programming competition at college level. This is a step up from high school level olympiads such as the IMO. Every year, participants from nearly 3000 universities and over 103 countries compete in solving real-world coding problems. This year's world finals took place in Baku, Azerbaijan on September 4, and brought together the top teams from earlier phases of the competition. Over a five-hour period, each team tackled a set of complex algorithmic problems. Final rankings hinged on two unforgiving principles: only perfect solutions earned points, and every minute counted. From the 139 competing teams, only the top four teams won gold medals. An advanced version of Gemini 2.5 Deep Think competed live in a remote online environment following ICPC rules, under the guidance of the competition organizers. It started 10 minutes after the human contestants and correctly solved 10 out of 12 problems, achieving gold-medal level performance under the same five-hour time constraint. See our solutions here. Gemini solved eight problems within just 45 minutes and two more problems within three hours, using a wide variety of advanced data structures and algorithms to generate its solutions. By solving 10 problems in a combined total time of 677 minutes, Gemini 2.5 Deep Think would be ranked in 2nd place overall, if compared with the university teams in the competition. Dr. Bill Poucher, ICPC Global Executive Director, stated: "The ICPC has always been about setting the highest standards in problem solving. Gemini successfully joining this arena, and achieving gold-level results, marks a key moment in defining the AI tools and academic standards needed for the next generation. Congratulations to Google DeepMind; this work will help us fuel a digital renaissance for the benefit of all."
[7]
OpenAI, Google reasoning models achieve gold-level scores in ICPC coding contest - SiliconANGLE
OpenAI, Google reasoning models achieve gold-level scores in ICPC coding contest OpenAI and Google LLC today disclosed that their latest reasoning models achieved gold-level performance in a recent coding competition. The ICPC, as the event is called, is the world's most prestigious college-level programming contest. It draws participants from about 3,000 universities. OpenAI says that its reasoning models achieved a perfect score, while Google's algorithm solved 10 of the 12 problems in this year's contest. The ICPC finals took place on September 4. The 139 participating teams had five hours to solve the dozen questions that the contest's organizers included in the test. Had the AI-generated submissions from OpenAI and Google been produced by humans, they would have won first and second place, respectively. Both companies' algorithms successfully solved Problem C, which none of the participating teams answered correctly. The task was to calculate the most efficient way of filling a set of reservoirs. "Problem C required finding a solution for distributing liquid through a network of interconnected ducts to a set of reservoirs, with the goal of finding a configuration of these ducts that fills all the reservoirs as quickly as possible," Google DeepMind researchers detailed in a blog post today. The company entered the contest with an "advanced" version of Gemini 2.5 Deep Think, a reasoning model it introduced in April. The algorithm tackles complex problems by generating a large number of potential answers in parallel. It then refines those answers and distills them into a single response. According to Google, the version of Gemini 2.5 Deep Think that participated in the ICPC used AI agents to generate multiple potential solutions to each problem. The agents had access to a terminal that enabled them to run and test code. After producing the initial code, they made refinements to improve the quality of the test responses. "Achieving gold-medal level at the ICPC has immediate, practical consequences for software development," the Google researchers wrote. "Beyond math and coding, our achievement demonstrates a powerful new capability in abstract reasoning. The skills needed for the ICPC -- understanding a complex problem, devising a multi-step logical plan and implementing it flawlessly -- are the same skills needed in many scientific and engineering fields." OpenAI participated in the ICPC with GPT-5 and an experimental reasoning model that isn't yet publicly available. GPT-5 solved 11 of the 12 problems with the reasoning model's help. The latter algorithm answered the last question, which was the most difficult of the set, on its own.
[8]
Google's Gemini AI achieves gold medal in prestigious ICPC coding competition, outperforming most human teams
Google has announced that its advanced AI model, Gemini 2.5 Deep Think, achieved a gold medal-level performance at the 2025 International Collegiate Programming Contest (ICPC) World Finals. The AI correctly solved 10 out of 12 complex coding problems, securing the second-highest overall score and outperforming the vast majority of human competitors. This achievement marks a significant step toward artificial general intelligence (AGI) and showcases the model's potential to revolutionize fields like software development and scientific research. The ICPC World Finals, held on September 4, 2025, in Baku, Azerbaijan, is considered the most prestigious university-level coding competition globally. Teams from nearly 3,000 universities were tasked with solving a series of intricate problems within a five-hour time limit, where only perfect solutions are accepted. Gemini 2.5 Deep Think operated as an automated team of AI agents, with multiple instances proposing, coding, and testing solutions collaboratively. This multi-agent approach allowed it to tackle complex challenges systematically. One of the most impressive feats was Gemini's solution to "Problem C," a complex optimization challenge that no human team managed to solve during the competition. The problem involved finding the optimal way to distribute liquid through a network of ducts. In under 30 minutes, Gemini developed a novel strategy using the minimax theorem, a concept from game theory, to find the solution. Google compared this moment to AlphaGo's famous "Move 37" in its 2016 match against Go champion Lee Sedol, where the AI made a creative and unexpected move that proved decisive. This achievement in competitive coding follows Gemini's success at the International Mathematical Olympiad in July 2025, where it also earned a gold medal. In its announcement, Google stated that these combined breakthroughs represent "a profound leap in abstract problem-solving -- marking a significant step on our path toward artificial general intelligence (AGI)." The company emphasized that the skills demonstrated by Gemini -- breaking down complex problems, creating multi-step logical plans, and executing them flawlessly -- are the same ones required for critical scientific and engineering tasks, such as designing new drugs or microchips. Google envisions a future where AI systems like Gemini collaborate with human experts, proposing unconventional ideas to accelerate scientific discovery and solve long-standing technical problems. While large language models have already had a significant impact on software development, Gemini's performance suggests a move toward more autonomous, reasoning-focused AI that can tackle real-world challenges with human-like ingenuity. Although the model failed to solve two problems that human teams did, its overall success points to a future where human-AI partnerships could drive innovation across industries.
[9]
Google's Gemini cracks problem no human could solve at global coding contest - The Economic Times
Google CEO Sundar Pichai has announced a major achievement by its advanced AI model, Gemini 2.5 Deep Think. The model delivered what he called a "gold-medal performance" at the 2025 International Collegiate Programming Contest (ICPC) World Finals, one of the world's top university-level programming competitions. Posting on X, Pichai wrote: "Incredible milestone: an advanced version of Gemini 2.5 Deep Think achieved gold-medal performance at the ICPC World Finals, a top global programming competition, solving an impressive 10/12 problems. Such a profound leap in abstract problem-solving - congrats to @googledeepmind!" Another notable achievement is that Gemini managed, in 30 minutes, to solve a problem no human team was able to solve during the entire competition. This comes just two months after the same model won a gold medal at the International Mathematical Olympiad (IMO). While competitions such as the IMO are aimed at high school students, the ICPC is an algorithmic programming contest at the university level. At the ICPC, participants from nearly 3,000 universities across more than 103 countries compete to solve real-world coding challenges. Competition details According to Google's blog post, Gemini 2.5 Deep Think took part in the ICPC competition in a live, remote setting, closely following official rules and under the supervision of the event organisers. The model began competing 10 minutes after the human teams and had the same five-hour time limit. Gemini quickly solved eight problems in just 45 minutes and completed two more within the next three hours. In total, it solved 10 out of 12 problems in 677 minutes of combined time, which would have placed it second overall if ranked alongside the university teams. Bill Poucher, ICPC Global Executive Director, commented: "Gemini successfully joining this arena, and achieving gold-level results, marks a key moment in defining the AI tools and academic standards needed for the next generation. Congratulations to Google DeepMind; this work will help us fuel a digital renaissance for the benefit of all." Explaining how Gemini reached this level of performance, Google wrote, "During the course of reinforcement learning, we trained Gemini to reason and generate code for some of the most difficult problems coders have faced, to learn from feedback on results and evolve its approaches." To solve problems, multiple Gemini agents suggest their own code, run tests, and then refine their answers based on all attempts. Google also claimed that its internal tests show this version of Gemini 2.5 Deep Think would have achieved gold-medal-level results at both the 2023 and 2024 ICPC World Finals, on par with the top 20 human coders globally. Finally, for everyday users, Google noted that a lighter version of Gemini 2.5 Deep Think is already available through the Gemini app for those with Google AI Ultra subscriptions.
[10]
World programming championship: How ChatGPT, Gemini and AI bots performed
AI systems redefine competitive programming, challenging human monopoly in algorithmic problem solving Baku, Azerbaijan, a city known as a cultural and historical crossroads, became the stage for a very different kind of history this September. At the 2025 International Collegiate Programming Contest (ICPC) World Finals, an event long celebrated for showcasing the brilliance of young human minds, the spotlight shifted in a dramatic new direction. For the first time, artificial intelligence systems stood side by side with elite student teams. And by the end of the five-hour contest, two AI competitors - OpenAI's GPT-5 and Google DeepMind's Gemini 2.5 Deep Think - had redefined what "competitive programming" even means. The ICPC has always symbolized intellectual grit: teams of three students from nearly 3,000 universities across 103 countries, racing to solve twelve algorithmic puzzles that test logic, math, and coding stamina. But this year, organizers introduced a groundbreaking twist, allowing AI systems to enter under strict supervision. Also read: OpenAI to build age-prediction system, restrict flirtation and suicide talk for teens Both OpenAI and DeepMind agreed to identical conditions. Their models received the same official problem set, submitted solutions through the standard judging portal, and faced real-time evaluation. No shortcuts, no hidden data, no tailored training. For five hours, GPT-5 and Gemini competed on the same terms as human teams, tasked with producing correct and efficient code for puzzles they had never seen before. This setup made the contest not just a competition, but an experiment: could general-purpose reasoning models rise to the level of the best human minds under live pressure? The most astonishing result came from OpenAI. Its ensemble of models, anchored by GPT-5, achieved a perfect score: 12 out of 12 problems solved. Eleven were accepted on the very first try. Only the toughest problem required multiple submissions, and even that fell after OpenAI deployed an experimental variant of its model. To put this in perspective, no human team in ICPC history has ever reached such perfection in this format. In Baku, the strongest student squads, from perennial powerhouses like Saint Petersburg State University and Tsinghua University, managed eleven. OpenAI's flawless sweep was more than a win. It was a signal that general AI has now crossed into territory once thought to be the exclusive domain of the human brain: adaptability, abstract reasoning, and creative algorithm design. Not to be overshadowed, Google's Gemini 2.5 delivered its own breathtaking performance. The system solved 10 out of 12 problems, which would have secured it a gold medal among human teams. More impressive was its blistering speed: Gemini cracked eight problems in just 45 minutes, a pace unmatched even by the fastest human competitors. Also read: Nothing to launch first AI-native devices in 2026, CEO hints they won't be phones But Gemini's crowning achievement was its handling of Problem C, a notoriously difficult systems optimization challenge. No human team solved it. Gemini did, by inventing a hybrid method that combined minimax game theory with dynamic programming, an approach many experts had never seen applied in this way. What makes the ICPC 2025 results so significant is not the raw numbers, but what they reveal about the nature of AI reasoning. These were not memorized solutions or brute-force calculations. The problems were novel, requiring abstract thinking, fresh modeling, and careful coding under time pressure. Both GPT-5 and Gemini showed the ability to reason on the fly, adapt strategies, and produce algorithms unfamiliar even to veteran engineers. The contest demonstrated that the age-old monopoly of humans over competitive programming had ended. The implications stretch far beyond ICPC. Competitive programming is more than sport, it has always been a training ground for innovation, shaping the engineers and scientists who power global tech. With AI systems now surpassing the best human teams, the field must grapple with new questions. Will future contests allow human-AI collaborations? Should there be separate AI divisions? How will education adapt when machines can outperform students not just in speed, but in originality? What is certain is that the ICPC 2025 will be remembered as a watershed moment. Human ingenuity, teamwork, and intuition remain irreplaceable - but the frontier of problem-solving has shifted. The algorithms are no longer just running with us. In many ways, they are already running ahead.
Share
Share
Copy Link
Google's Gemini 2.5 and OpenAI's GPT-5 achieve gold medal performances in the International Collegiate Programming Contest (ICPC) World Finals, solving complex problems that stumped human competitors. This breakthrough demonstrates AI's advanced reasoning capabilities and potential impact on scientific discovery.
In a groundbreaking development, artificial intelligence models from Google DeepMind and OpenAI have achieved remarkable success in the 2025 International Collegiate Programming Contest (ICPC) World Finals, often dubbed the 'coding Olympics'
1
2
3
. This milestone highlights the rapid advancement of AI technology and its increasing capability in complex problem-solving scenarios.Source: Digit
Google's Gemini 2.5 Deep Think, an advanced AI model, secured a gold medal-level performance by correctly solving 10 out of 12 problems
1
2
. This placed the AI model second overall, surpassing 135 of 139 human teams1
. Notably, Gemini solved a problem that eluded all human competitors, showcasing its superior problem-solving prowess1
3
.Source: Economic Times
OpenAI reported that their GPT-5 model would have achieved a perfect score, solving all 12 problems in the competition, had it officially competed
4
5
. This hypothetical feat would have outpaced even the top human teams, none of whom managed to solve every problem5
.The ICPC World Finals is a prestigious global event where university teams tackle 12 complex programming problems within a five-hour limit
2
3
. The AI models exhibited significant advantages:Source: Google DeepMind
1
5
.1
2
.1
4
.Related Stories
This achievement is widely seen as a substantial stride towards Artificial General Intelligence (AGI)
2
3
. The AI models' ability to excel in tasks demanding abstract reasoning, creativity, and flawless execution suggests immense potential for scientific and engineering applications2
3
.Google DeepMind's vice president, Quoc Le, described this as "a historic moment towards AGI," underscoring the potential for these systems to drive breakthroughs in fields like drug design and semiconductor development
3
5
.While impressive, experts advise caution against broad generalizations about AI's ability to write production-ready software
3
. The competition prioritizes speed under pressure, differing from real-world development3
.Nevertheless, this rapid progress indicates a future where AI systems could significantly augment human intelligence, offering original solutions to complex problems across diverse disciplines
3
5
. The potential for AI to resolve previously intractable challenges and accelerate scientific discovery is becoming increasingly clear1
2
3
4
5
.Summarized by
Navi
[3]
20 Jul 2025•Science and Research
26 Jul 2024
21 Jul 2025•Technology