Ask five people which country is winning at artificial intelligence and you’ll probably get five answers. One will point to the number of frontier models released, another to robot factories, and a third to government readiness. The answer shifts depending on what you count.
This article leans on three public yardsticks: Stanford’s AI Index, Stanford’s Global AI Vibrancy Tool, and the Oxford Insights Government AI Readiness Index. Each measures something different, so a country can shine on one list and look ordinary on another. The seven below show up near the top across them, in different ways.
1. United States: Still the Pace Setter
Stanford’s 2026 AI Index counts 59 notable US models for 2025, against 35 for China and 8 for South Korea.[1] The year before, US institutions produced 40 notable models, well ahead of China’s 15 and Europe’s combined three.[2] Those counts come from Epoch AI, which defines “notable” by things like state of the art results, citations and wide use.
The broader scorecards agree. On Stanford’s 2025 Global AI Vibrancy Tool the US topped the list with 78.60, with China second at 36.95.[3] The US also leads the Oxford Insights readiness ranking, with an overall score of 88.36.[4]
2. China: Closing the Gap Fast
China leads in publication volume, citations and patent grants, and its share of the world’s 100 most cited AI papers rose from 33 in 2021 to 41 in 2024.[5] It also dominates physical automation. China installed 295,000 industrial robots in 2024, while Japan installed roughly 44,500 and the United States 34,200.[6]
On model quality, the distance has shrunk. US and Chinese models have traded the top spot multiple times since early 2025, with DeepSeek R1 briefly matching the leading US model in February 2025.[5] In the Oxford Insights readiness index, China now ranks 6th, up from 23rd in 2024.[7]
3. India: The Fastest Climber
Stanford’s 2025 Vibrancy Tool put India third, behind only the US and China, up from 7th in 2023.[8] India’s score was 21.59.[9] That’s a long way behind the leaders, though it sits comfortably ahead of several wealthy economies.
The rise has specific causes. It was driven by research output, startups and AI talent, though India slipped in policy and governance.[8] Coverage of the ranking also stresses the gap India still has to bridge with the two leaders.[3]
4. South Korea: Small Country, Big Patent Footprint
South Korea ranked fourth on the 2025 Vibrancy Tool with a score of 17.24.[3] In model building it was third in 2025, with eight notable models, a distant third behind the US and China but well ahead of every other country. Summaries of the 2026 AI Index also name South Korea as the leader in AI patents per capita.[1]
Policy has kept pace. South Korea hosted the most recent global AI summit in 2024.[10] In the 2024 Oxford Insights readiness edition, it sat second only to Singapore among the global top three, with a score of 79.98.[11]
5. United Kingdom: Strong on Governance, Lighter on Models
The UK placed fifth on the 2025 Vibrancy Tool with a score of 16.64.[12][9] It took a global leadership role by hosting the world’s first AI safety summit in 2023.[10] Its readiness is also strong. In the 2024 Oxford Insights edition, the UK scored 78.88, narrowly behind France in Western Europe.[11]
Raw model output is the weaker spot. Epoch’s database credits the UK with just one notable model in 2025.[13] That shows how readiness and research strength can pull in different directions.
6. France: Europe’s Most Visible Model Builder
France has the best readiness result in Europe. It ranks 2nd globally in the 2025 Oxford Insights index with a score of 80.81.[14] In June 2025 its National Digital Council became the National Council for AI and Digital.[14]
It also has a homegrown lab with real output. Mistral AI released Mistral Large 3 on 2 December 2025, with 675 billion total parameters, under an Apache 2.0 licence.[15] Still, Europe as a whole is thin at the frontier, because the AI Index says it produced only two notable models in 2025.[15]
7. Singapore: Government Readiness in a Compact Package
In the 2024 Oxford Insights readiness index, Singapore led the world with 84.25.[11] A summary of the 2025 edition has it 7th, with strength in government driven AI adoption.[16] Those two results suggest the country has stayed near the top even as the index methodology changed.
Singapore also placed sixth on the 2025 Vibrancy Tool with a score of 16.43.[12] The city state doesn’t have the population to compete on volume. What it offers instead is policy, infrastructure and an environment that’s friendly to AI deployment.
Falling behind looks different, and the data point mostly to one region. Sub Saharan Africa scores 29.12 on the 2025 Oxford Insights readiness index, against 81.51 for North America.[17] The indexes don’t draw a neat line at ten countries, but the weakest results cluster in the same place.
On AI infrastructure, Burundi, Chad, Niger and the Republic of Congo each score below 20.[18] Half of Sub Saharan African countries score below 35 on that measure.[18] Researchers describe Somalia as ranking at the bottom of global AI readiness indices.[19]
No country in the region scores above 56 out of 100 on infrastructure, compared with 89.27 for the United States and 76.92 for China.[18] By the end of 2025, South Africa was the only African country with an AI adoption rate above one fifth, compared with 64 percent in the UAE.[20] African nations account for roughly one percent of the world’s AI computing capacity.[21]
Ambition isn’t absent. Sixteen countries have national AI strategies, and the African Union adopted a Continental AI Strategy in July 2024.[17] Africa’s AI market is projected to grow from 4.5 billion US dollars in 2025 to 16.5 billion by 2030.[17]
