iTechGuides is reader-supported. When you buy through links on our site, we may earn an affiliate commission. As an Amazon Associate I earn from qualifying purchases. Learn more
Neither country wins the AI race outright. Which one leads depends on the measure you choose and the date it was taken. In Stanford HAI’s 2026 AI Index, the United States holds the edge in private AI investment, notable model production, higher-impact patents and data-center count. China leads in research publication volume, citations, patent grants and industrial robot installations. At the frontier of model performance, the two countries have traded the lead repeatedly since early 2025, so a single winner is the wrong way to frame the answer.
The scorecard by measure
The table below sets out each measure the four charts use, along with the leading country and the figure reported. Where the source does not give a figure, the cell says so.
| Measure | Leader | Figure and date | Source |
|---|---|---|---|
| Frontier model performance | Near parity; lead has changed hands | Top U.S. model ahead of top Chinese model by 2.7% as of March 2026 | Stanford HAI 2026 AI Index |
| Notable AI models produced | United States | 59 U.S. vs. 35 Chinese, 2025 | Stanford HAI 2026 AI Index, using an Epoch AI dataset |
| Private AI investment | United States | $285.9 billion U.S. vs. $12.4 billion China, 2025 | Stanford HAI 2026 AI Index |
| Research publication volume | China | Not stated | Stanford HAI 2026 AI Index |
| Citations | China | Not stated | Stanford HAI 2026 AI Index |
| Patent grants | China | Not stated | Stanford HAI 2026 AI Index |
| Higher-impact patents | United States | Not stated | Stanford HAI 2026 AI Index |
| Industrial robot installations | China | Not stated | Stanford HAI 2026 AI Index |
| Data centers | United States | 5,427, more than ten times any other country | Stanford HAI 2026 AI Index |
Chart 1: Frontier model performance over time
This chart tracks which country’s best model is ahead, and that position has changed more than once. Stanford HAI reports that U.S. and Chinese frontier models have traded the lead several times since early 2025. In February 2025, DeepSeek-R1 briefly matched the top U.S. model.
Recommended Free Tools
The figure to label is the March 2026 snapshot. On the report’s comparison, the top U.S. model led the top Chinese model by 2.7%. The report attributes that U.S. model to Anthropic. Stanford HAI’s own summary of the position reads: “The U.S.-China AI model performance gap has effectively closed.”
#1 Best Overall
Draw the lead as a line that switches sides, not as a trend. A 2.7% margin on one comparison at one date is neither a permanent advantage nor a measure of overall AI capability.
Chart 2: Notable model production
For 2025, Stanford HAI counts 59 notable U.S. models and 35 notable Chinese models. The count draws on an Epoch AI dataset that is curated by hand. A model qualifies on criteria such as advancing the state of the art, historical significance or a high citation count.
This makes the chart a count of notable models, not of all models released. Label it that way. A single curation decision can shift totals, and a model that is widely used but does not meet those criteria will not appear.
Chart 3: Private AI investment
Stanford HAI reports $285.9 billion in private AI investment in the United States in 2025, against $12.4 billion in China. On this measure, the U.S. figure is roughly 23 times the Chinese one.
Rank #3
- Incredibly Light. Surprisingly Thin. - LG gram is designed to go wherever you do. Weighing just 2.5 lbs. with an ultra-slim 0.7-inch profile, it slips easily into your bag and feels light in hand—making it effortless to carry, commute, and work from anywhere.
- Remarkably Light. Reliably Strong. - LG gram has passed seven military-grade durability tests, striking an impressive balance between a highly portable, lightweight metal build and the confidence to handle everyday movement and travel.
- Power That Last with Smart Efficiency - LG gram combines a high-capacity 72Wh battery with AI-driven power management to optimize efficiency based on your usage. The result is up to 32 hours of video playback for} long-lasting performance that keeps up with your day—at home, at work, or wherever you go.
- AMD Ryzen AI Performance - Powered by AMD’s AI-optimized Ryzen processor with Radeon Graphics and a built-in NPU, LG gram delivers smooth multitasking and responsive performance. Fast 32GB LPDDR5x memory and 1TB NVMe storage keep everything moving without slowdowns.
- Dual AI for Always-On Intelligence - LG gram’s Dual AI—powered by EXAONE 3.5, LG’s AI solution—combines gram chat On-Device AI and gram chat Cloud AI to deliver seamless assistance. gram chat On-Device AI enables fast document search and summarization directly on your PC, while gram chat Cloud AI expands capabilities when connected—so everyday tasks stay smooth, responsive, and uninterrupted.
The caveat belongs on the chart itself. The figures count private money only. China’s government guidance funds, which steer state capital toward favored sectors, are not captured on the same basis. The comparison therefore likely understates China’s total AI spending, and it should not be presented as a measure of national AI spending.
Chart 4: Research output versus research impact
Split this chart into two panels, because the measures point in different directions.
- Output and reach (China leads): publication volume, citations and patent grants.
- Impact (United States leads): higher-impact patents, the report’s measure of patent influence.
Keep these as separate bars. More papers and more granted patents do not automatically mean more influential work, and a combined score would hide that distinction.
Beyond the four charts: infrastructure and talent
Infrastructure and talent give context, but they do not fit neatly into a China-versus-U.S. chart, so treat them as sidebars.
Best Value
Data centers and chips
Stanford HAI counts 5,427 data centers in the United States, more than ten times the number in any other country. Almost every leading AI chip is fabricated by TSMC in Taiwan, and TSMC’s U.S. expansion began operating in 2025. Together these facts show both scale and dependence: the U.S. has the largest data-center footprint, while leading-chip manufacturing is concentrated in one company. They are not a direct comparison of Chinese and U.S. data-center capacity.
Talent flows: three ways to count
Carnegie Endowment’s 2026 talent analysis uses a sample of elite AI researchers from the 2025 NeurIPS conference author cohort. Its figures answer different questions depending on how talent is counted.
- Origin: 57% of the sampled talent originated in China and 13% in the United States, based on where people earned their undergraduate degrees.
- Net migration: in 2025, the analysis shows a net gain of 2,145 researchers for the United States and a net loss of 1,729 for China.
- Current workplace: a third lens asks where people work now. The origin and migration figures above do not answer that question.
The NeurIPS cohort is a useful proxy for elite research talent, but it is one conference, not the whole AI workforce.
Matching the measure to your question
Choose the chart that fits the question you are asking.
Quick Recap
| Your question | Measure to use | Where it is covered |
|---|---|---|
| Which country’s best model is ahead right now? | Frontier model comparison, March 2026 | Chart 1 |
| Which country is producing more notable models? | Notable model counts, 2025 | Chart 2 |
| Where is private AI money going? | Private AI investment, 2025 | Chart 3 |
| Who publishes more, and whose work is cited more? | Publication volume and citations, shown separately | Chart 4 |
| Whose patents have more influence? | Higher-impact patents, distinct from patent grants | Chart 4 |
| Who has the physical computing base? | Data-center count and chip fabrication dependence | Beyond the four charts |
| Who is drawing elite AI researchers? | Talent origin and net migration, which answer different questions | Beyond the four charts |
Product prices and availability are accurate as of the date/time indicated and are subject to change. Any price and availability information displayed on Amazon at the time of purchase will apply.
