{"id":4436,"date":"2026-09-13T10:56:47","date_gmt":"2026-09-13T10:56:47","guid":{"rendered":"https:\/\/www.vaultinsider.top\/?p=4436"},"modified":"2026-09-13T10:56:47","modified_gmt":"2026-09-13T10:56:47","slug":"chinese-ai-labs-are-tightening-the-gap-with-the-us-by-just-being-more-efficient","status":"publish","type":"post","link":"https:\/\/www.vaultinsider.top\/?p=4436","title":{"rendered":"Chinese AI labs are tightening the gap with the US by just being more efficient"},"content":{"rendered":"

<\/p>\n

The race between the world\u2019s two biggest economies to dominate AI recently escalated with new accusations and raised the question of how China came to rival the U.S.\u2019s AI capabilities.\u00a0<\/p>\n

\n

Six Chinese AI companies, U.S. officials alleged earlier this week, found a shortcut to closing the gap with American AI labs by buying bulk subscriptions to their American rivals and then training on the outputs. <\/p>\n

The FBI, the National Security Agency, and the Cybersecurity and Infrastructure Security Agency said Tuesday that DeepSeek, Moonshot, and four others extracted \u201ccapabilities worth billions\u201d in this way since 2024, a method the agencies claimed let DeepSeek understate its $5.6 million training cost.<\/p>\n

China\u2019s foreign affairs ministry said the country\u2019s AI development \u201cis a result of high-level scientific and technological self-reliance\u201d and called the accusations \u201cgroundless.\u201d<\/p>\n

The allegations are one potential explanation for why Chinese AI models are estimated to be neck-and-neck with their U.S. counterparts, with a report from Stanford putting Anthropic\u2019s top model ahead of DeepSeek\u2019s by just 2.7% earlier this year.\u00a0<\/p>\n

But analysts say there\u2019s a different advantage Chinese labs developed to compete: squeezing out more value from limited resources.\u00a0<\/p>\n

More value for fewer tokens<\/strong><\/h2>\n

There\u2019s a technique Chinese labs perfected that stems from \u201cattention,\u201d the mechanism Google researchers introduced in a landmark 2017 paper that underlies every large language model.<\/p>\n

Attention is what lets a model look at each piece of text in relation to all the others and decide which connections matter most. That helps it understand context, but it gets computationally more expensive the longer the context window gets.\u00a0<\/p>\n

Brendan Burke, the semiconductors and supply chain analyst at tech research firm Futurum Group, told Fortune<\/em> that Chinese labs figured out a shortcut to make the attention mechanism cheaper and more efficient.\u00a0<\/p>\n

\u201cChinese labs found algorithms that reduce the complexity of those calculations by an order of magnitude, and then achieve better results because they\u2019re able to summarize the most relevant tokens,\u201d he said.<\/p>\n

Necessity was the mother of invention, as the U.S. restricted China\u2019s access to Nvidia\u2019s best chips, pushing it toward domestic alternatives like Huawei, which limited China\u2019s access to the highest-performing compute available in the U.S. Meanwhile, the U.S. has 74% of the world\u2019s compute, according to a White House report, and it also has hyperscalers pouring billions into generating more through data centers.\u00a0<\/p>\n

\u201cBecause they had less compute to work with, they found that computationally efficient method instead of just throwing more compute at an inefficient technique, as U.S. labs initially did,\u201d Burke said about China.\u00a0<\/p>\n

By contrast, U.S. frontier labs can be \u201ctoken hogs\u201d because their AI systems are designed to be exploratory and to test base models, he explained.<\/p>\n

The trade-off between the two can be quantified. Ameya Kanitkar, cofounder of the AI measurement platform Larridin, told Fortune<\/em> that in the enterprise workflows Larridin tracks, Chinese models like GLM 5.2 and Kimi 2.6 and 2.7 handle around 75% of engineering tasks \u201creasonably well\u201d at a fifth of the cost of the U.S. ones.<\/p>\n

\u201cFrontier U.S. models still have an advantage on the most complex tasks, but Chinese open-weight models are becoming more than capable enough for the majority of everyday enterprise engineering work,\u201d Kanitkar said.\u00a0\u00a0<\/p>\n

The cost difference might end up mattering as AI spending consumes a bigger share of corporate budgets, with 20% of business leaders surveyed by McKinsey saying AI-related costs like buying tokens is constraining their use of it.\u00a0<\/p>\n

U.S. enterprises warming up to Chinese models<\/strong><\/h2>\n

Cost-efficiency isn\u2019t the only appeal. DeepSeek\u2019s R1 reasoning model was made available for download through platforms like Hugging Face, allowing companies to run and adapt versions of the model themselves. This helps companies by giving them an open-source model to work with and fine-tune to meet their needs without depending on a closed model, with the option to run it through a U.S.-based cloud provider like Amazon Web Services.\u00a0<\/p>\n

This flexibility helped U.S. businesses that might have been wary of sending data to a China-based company warm up to DeepSeek, and this applies to other Chinese models as well. Hugging Face reported Chinese open-source models accounted for 41% of total downloads last year, a larger share than U.S. ones.\u00a0\u00a0<\/p>\n

DoorDash CEO Andy Fang said using Moonshot AI\u2019s Kimi was \u201ccheaper\u201d and \u201cbetter quality\u201d without degrading the quality of code. AI coding startup Cursor also used Kimi to help build its Composer 2 coding agent. Airbnb and Siemens are also experimenting with Alibaba and DeepSeek models, with AirbnB CEO Brian Chesky calling Qwen \u201cfast and cheap.\u201d<\/p>\n

Chinese models are even starting to replace their American counterparts in more specialized work. Thomson Reuters said it built an in-house model called Thomson-1 by adapting Alibaba\u2019s open-source Qwen model to handle document-review work previously done by Claude.<\/p>\n

Data suggests the enterprise shift is becoming more visible. Ramp\u2019s AI index showed that the share of businesses paying for platforms with access to open-source and Chinese-developed models rose to 6.1% of total AI-spending businesses in July from 4.5% in January.\u00a0<\/p>\n

But companies experimenting with Chinese models doesn\u2019t necessarily mean they\u2019re leaving behind U.S. ones, which are still months ahead in performance. Mike Finley, chief technology officer of enterprise AI analytics firm AnswerRocket, told Fortune<\/em> the output of U.S. AI companies still serves as the \u201cexistence proof\u201d for Chinese labs to innovate off of.\u00a0<\/p>\n

\u201cThe work they do would simply not be possible without the frontier labs blazing the trail,\u201d Finley said.\u00a0\u00a0<\/p>\n<\/div>\n

#Chinese #labs #tightening #gap #efficient<\/p>\n","protected":false},"excerpt":{"rendered":"

The race between the world\u2019s t… <\/p>\n","protected":false},"author":1,"featured_media":4437,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[173,657,7425,1697,811,385,7424],"class_list":["post-4436","post","type-post","status-publish","format-standard","has-post-thumbnail","hentry","category-finance-news","tag-china","tag-chinese","tag-efficient","tag-gap","tag-labs","tag-tech","tag-tightening"],"yoast_head":"\nChinese AI labs are tightening the gap with the US by just being more efficient - Finance News<\/title>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/www.vaultinsider.top\/?p=4436\" \/>\n<meta property=\"og:locale\" content=\"zh_CN\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Chinese AI labs are tightening the gap with the US by just being more efficient - Finance News\" \/>\n<meta property=\"og:description\" content=\"The race between the world\u2019s t…\" \/>\n<meta property=\"og:url\" content=\"https:\/\/www.vaultinsider.top\/?p=4436\" \/>\n<meta property=\"og:site_name\" content=\"Finance News\" \/>\n<meta property=\"article:published_time\" content=\"2026-09-13T10:56:47+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg?w=2048\" \/>\n<meta name=\"author\" content=\"admin\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u4f5c\u8005\" \/>\n\t<meta name=\"twitter:data1\" content=\"admin\" \/>\n\t<meta name=\"twitter:label2\" content=\"\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4\" \/>\n\t<meta name=\"twitter:data2\" content=\"5 \u5206\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"Article\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#article\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436\"},\"author\":{\"name\":\"admin\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/#\\\/schema\\\/person\\\/ed40724a3e5d482ce66781a042c3c31e\"},\"headline\":\"Chinese AI labs are tightening the gap with the US by just being more efficient\",\"datePublished\":\"2026-09-13T10:56:47+00:00\",\"mainEntityOfPage\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436\"},\"wordCount\":943,\"commentCount\":0,\"image\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.vaultinsider.top\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/GettyImages-2276117405-e1789168021100.jpg\",\"keywords\":[\"China\",\"Chinese\",\"efficient\",\"gap\",\"labs\",\"Tech\",\"tightening\"],\"articleSection\":[\"Finance News\"],\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"CommentAction\",\"name\":\"Comment\",\"target\":[\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#respond\"]}]},{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436\",\"url\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436\",\"name\":\"Chinese AI labs are tightening the gap with the US by just being more efficient - Finance News\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/www.vaultinsider.top\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/GettyImages-2276117405-e1789168021100.jpg\",\"datePublished\":\"2026-09-13T10:56:47+00:00\",\"author\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/#\\\/schema\\\/person\\\/ed40724a3e5d482ce66781a042c3c31e\"},\"breadcrumb\":{\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#breadcrumb\"},\"inLanguage\":\"zh-Hans\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"zh-Hans\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#primaryimage\",\"url\":\"https:\\\/\\\/www.vaultinsider.top\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/GettyImages-2276117405-e1789168021100.jpg\",\"contentUrl\":\"https:\\\/\\\/www.vaultinsider.top\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/GettyImages-2276117405-e1789168021100.jpg\",\"width\":1200,\"height\":600},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/?p=4436#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"\u9996\u9875\",\"item\":\"https:\\\/\\\/www.vaultinsider.top\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Chinese AI labs are tightening the gap with the US by just being more efficient\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/#website\",\"url\":\"https:\\\/\\\/www.vaultinsider.top\\\/\",\"name\":\"Finance News\",\"description\":\"Just another Finance News site\",\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/www.vaultinsider.top\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"zh-Hans\"},{\"@type\":\"Person\",\"@id\":\"https:\\\/\\\/www.vaultinsider.top\\\/#\\\/schema\\\/person\\\/ed40724a3e5d482ce66781a042c3c31e\",\"name\":\"admin\",\"url\":\"https:\\\/\\\/www.vaultinsider.top\\\/?author=1\"}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"Chinese AI labs are tightening the gap with the US by just being more efficient - Finance News","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/www.vaultinsider.top\/?p=4436","og_locale":"zh_CN","og_type":"article","og_title":"Chinese AI labs are tightening the gap with the US by just being more efficient - Finance News","og_description":"The race between the world\u2019s t…","og_url":"https:\/\/www.vaultinsider.top\/?p=4436","og_site_name":"Finance News","article_published_time":"2026-09-13T10:56:47+00:00","og_image":[{"url":"https:\/\/fortune.com\/img-assets\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg?w=2048","type":"","width":"","height":""}],"author":"admin","twitter_card":"summary_large_image","twitter_misc":{"\u4f5c\u8005":"admin","\u9884\u8ba1\u9605\u8bfb\u65f6\u95f4":"5 \u5206"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"Article","@id":"https:\/\/www.vaultinsider.top\/?p=4436#article","isPartOf":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436"},"author":{"name":"admin","@id":"https:\/\/www.vaultinsider.top\/#\/schema\/person\/ed40724a3e5d482ce66781a042c3c31e"},"headline":"Chinese AI labs are tightening the gap with the US by just being more efficient","datePublished":"2026-09-13T10:56:47+00:00","mainEntityOfPage":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436"},"wordCount":943,"commentCount":0,"image":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436#primaryimage"},"thumbnailUrl":"https:\/\/www.vaultinsider.top\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg","keywords":["China","Chinese","efficient","gap","labs","Tech","tightening"],"articleSection":["Finance News"],"inLanguage":"zh-Hans","potentialAction":[{"@type":"CommentAction","name":"Comment","target":["https:\/\/www.vaultinsider.top\/?p=4436#respond"]}]},{"@type":"WebPage","@id":"https:\/\/www.vaultinsider.top\/?p=4436","url":"https:\/\/www.vaultinsider.top\/?p=4436","name":"Chinese AI labs are tightening the gap with the US by just being more efficient - Finance News","isPartOf":{"@id":"https:\/\/www.vaultinsider.top\/#website"},"primaryImageOfPage":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436#primaryimage"},"image":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436#primaryimage"},"thumbnailUrl":"https:\/\/www.vaultinsider.top\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg","datePublished":"2026-09-13T10:56:47+00:00","author":{"@id":"https:\/\/www.vaultinsider.top\/#\/schema\/person\/ed40724a3e5d482ce66781a042c3c31e"},"breadcrumb":{"@id":"https:\/\/www.vaultinsider.top\/?p=4436#breadcrumb"},"inLanguage":"zh-Hans","potentialAction":[{"@type":"ReadAction","target":["https:\/\/www.vaultinsider.top\/?p=4436"]}]},{"@type":"ImageObject","inLanguage":"zh-Hans","@id":"https:\/\/www.vaultinsider.top\/?p=4436#primaryimage","url":"https:\/\/www.vaultinsider.top\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg","contentUrl":"https:\/\/www.vaultinsider.top\/wp-content\/uploads\/2026\/09\/GettyImages-2276117405-e1789168021100.jpg","width":1200,"height":600},{"@type":"BreadcrumbList","@id":"https:\/\/www.vaultinsider.top\/?p=4436#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"\u9996\u9875","item":"https:\/\/www.vaultinsider.top\/"},{"@type":"ListItem","position":2,"name":"Chinese AI labs are tightening the gap with the US by just being more efficient"}]},{"@type":"WebSite","@id":"https:\/\/www.vaultinsider.top\/#website","url":"https:\/\/www.vaultinsider.top\/","name":"Finance News","description":"Just another Finance News site","potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/www.vaultinsider.top\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"zh-Hans"},{"@type":"Person","@id":"https:\/\/www.vaultinsider.top\/#\/schema\/person\/ed40724a3e5d482ce66781a042c3c31e","name":"admin","url":"https:\/\/www.vaultinsider.top\/?author=1"}]}},"_links":{"self":[{"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/posts\/4436","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=4436"}],"version-history":[{"count":0,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/posts\/4436\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=\/wp\/v2\/media\/4437"}],"wp:attachment":[{"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=4436"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=4436"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.vaultinsider.top\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=4436"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}