ChinaBizInsight

NEW NEW THREE 2026 H1 Update

Beyond the “New Three”

Robotics, AI, and Biotech Emerge as China’s Next Export Powerhouses

📅 August 16, 2026 ⏱ 12 min read 🏷️ #AI #Robotics #Biotech #NewNewThree #ChinaExport

Executive Summary: Just two years ago, the world was talking about China’s “New Three” exports—NEVs, lithium batteries, and solar PV. In 2026, the conversation has shifted. Robotics, artificial intelligence, and innovative biotech—what observers are calling the “New New Three”—have emerged as China’s next export powerhouses. In the first half of 2026, China’s AI-related product exports surpassed $480 billion, up 47.3% year-on-year. Industrial robot exports reached 6.29 billion yuan, sold to 141 countries. And Chinese biotech out-licensing deals hit $110 billion, with Chinese companies claiming 8 of the top 10 global biopharma transactions. This report breaks down the numbers, the drivers, the出海 logic, and what it all means for overseas investors, technology partners, and buyers.

1. The “Three Jumps”: From Old to New New

Understanding China’s export evolution requires looking at three distinct generations of products—what analysts call the “three jumps” of Chinese manufacturing[reference:0].

Generation Representative Products Driving Force Era
“Old Three”
老三样
👕 Apparel · 🪑 Furniture · 📺 Home Appliances Cost advantage 1980s–2000s
“New Three”
新三样
🚗 NEVs · 🔋 Lithium Batteries · ☀️ Solar PV Industrial chain & green transition 2020s–present
“New New Three”
新新三样
🤖 Robotics · 🧠 AI · 💊 Innovative Biotech Original innovation Emerging now

As one commentator put it: “The first two jumps jumped on cost and manufacturing; this jump jumps on technology itself“[reference:1]. The shift from “Old Three” to “New Three” to “New New Three” marks a fundamental transformation—from selling things you can hold in your hands to selling capabilities you can’t see[reference:2].

Tech leaders have taken note. At a recent policy briefing, China’s Minister of Science and Technology noted that AI, robotics, and biotech are developing rapidly and are expected to become new pillar industries for the country[reference:3]. This isn’t just government rhetoric—the numbers tell the story.

2. Artificial Intelligence: The Global Digital Backbone

If there’s one sector that embodies the “New New Three” revolution, it’s AI. The numbers are staggering—and they’re reshaping how the world thinks about Chinese technology.

📊 AI by the Numbers (H1 2026)

$480B+ — AI-related product exports, up 47.3%[reference:4][reference:5]
36.11T — Weekly token calls (July 2026), #1 globally[reference:6]
12+ — Consecutive weeks ranked #1 globally[reference:7]
Top 6 — All six monthly LLM rankings held by Chinese models[reference:8]
63.5% — Chinese models’ share of global weekly token volume[reference:9]
100B+ — Cumulative downloads of Chinese open-source models[reference:10]
41% — Chinese open-source model downloads on Hugging Face[reference:11]
2.7% — Performance gap between Chinese & US frontier models[reference:12]

Let’s unpack what these numbers mean.

🌐 Global Token Dominance

According to data from OpenRouter, the world’s largest AI model API aggregation platform, Chinese AI models have held the #1 position in weekly token calls for 12 consecutive weeks as of July 2026[reference:13]. In the week of July 27 to August 2, 2026, the top five models by weekly token volume were all Chinese: DeepSeek V4 Flash (7.22T), Xiaomi MiMo-V2.5 (6.3T), Tencent HY3 (4.82T), DeepSeek V4 Pro (3.28T), and Zhipu GLM-5.2 (2.89T)[reference:14]. Chinese models accounted for 63.5% of global weekly token volume, compared to 35.5% for US models[reference:15].

In July 2026, the monthly LLM ranking told the same story: the top six spots were all held by Chinese open-source models—Xiaomi MiMo-V2.5, DeepSeek V4 Flash, Tencent HY3, MiniMax M3, Zhipu GLM-5.2, and DeepSeek V4 Pro[reference:16].

What’s driving this? Cost and performance. In early August 2026, DeepSeek V4 Flash launched with an API pricing of 1 yuan per million input tokens and 2 yuan per million output tokens—with a 98% cache hit discount that can reduce costs to as low as 0.02 yuan per million tokens for repeated calls[reference:17]. OpenAI responded by cutting prices 80% on the same day—but still couldn’t catch up[reference:18]. As one developer put it: “We want DeepSeek’s price”[reference:19].

In the week of August 3–9, 2026, global token calls reached 69 trillion, up 21.48% week-over-week. China’s share was 34.25 trillion tokens, up 21.76%—marking 15 consecutive weeks of Chinese models leading the US[reference:20].

📥 Open-Source Ecosystem Leadership

Beyond commercial API calls, China’s open-source AI ecosystem has achieved global leadership. According to Hugging Face’s Spring 2026 report, Chinese-developed open-source models accounted for 41% of the platform’s downloads—surpassing the US at 36.5% for the first time[reference:21]. Cumulative downloads of Chinese open-source models have exceeded 10 billion[reference:22][reference:23]. “One in every 10 large model downloads globally is now from China,” as one observer noted[reference:24].

Chinese AI is no longer just a domestic phenomenon—it’s a global digital infrastructure. From AI classrooms in Kazakhstan to water management systems in Thailand, Chinese models are being integrated into real-world applications worldwide[reference:25][reference:26].

As a Customs official put it: “Chinese manufacturing precisely matches global demand”—and nowhere is that more evident than in AI[reference:27].

3. Robotics: From Assembly Lines to Operating Rooms

Chinese robotics has achieved what many thought impossible just a few years ago: in 2025, China became a net exporter of industrial robots for the first time[reference:28]. In 2026, the sector has only accelerated.

🤖 Robotics by the Numbers (H1 2026)

12.947M — Total robot units exported[reference:29]
24.85B — Total robot export value (yuan)[reference:30]
6.29B — Industrial robot exports (yuan), +18.6%[reference:31][reference:32]
141 — Countries receiving industrial robots[reference:33]
480M — Surgical robot exports (yuan), +3.3x[reference:34][reference:35]
49 — Countries receiving surgical robots (up from 23)[reference:36]
10,000+ — Smart bionic robots exported[reference:37][reference:38]
90+ — Countries receiving smart bionic robots[reference:39]
18.09B — Cleaning + bionic robot exports (yuan)[reference:40]

🏭 Industrial Robots: The Workhorse

Industrial robot exports reached 6.29 billion yuan in H1 2026, up 18.6%, sold to 141 countries and regions[reference:41][reference:42]. Chinese robots are now fixtures on factory floors worldwide—from Renault’s plant in Brazil to Proton’s factory in Malaysia[reference:43]. As one commentary noted, “Robot exports aren’t just a piece of equipment—they’re a complete intelligent control system and maintenance solution”[reference:44][reference:45]. Technical teams can now complete trajectory simulations in virtual worlds without waiting for equipment to ship overseas[reference:46].

🏥 Surgical Robots: The Growth Leader

Surgical robot exports hit 480 million yuan in H1 2026, up 3.3 times year-on-year[reference:47][reference:48]. Even more impressive: the number of export markets expanded from 23 to 49 countries[reference:49]. Chinese surgical robots are gaining international recognition for their precision and reliability, with Customs officials noting that “Chinese intelligent medical devices are benefiting more and more patients worldwide”[reference:50].

🧠 Smart Bionic Robots: The New Frontier

AI-integrated smart bionic robots—a category that didn’t even have its own customs code a year ago—saw exports exceed 10,000 units in H1 2026, reaching over 90 countries and regions[reference:51][reference:52]. Smart bionic robots and cleaning robots together accounted for 18.09 billion yuan in exports[reference:53]. These robots are being deployed in diverse scenarios: equipment inspection, scientific research, public services, and more[reference:54]. Chinese welding and搬运 robots saw export growth exceeding 60%[reference:55].

At the 2026 CES in the US, Chinese companies occupied more than half of the humanoid robot exhibition area[reference:56]. Unitree Robotics, with its quadruped and bipedal humanoid robots, has become a “new favorite” in global research and education markets[reference:57].

Chinese robotics companies are also building production capacity overseas. Chinese firms have established humanoid robot production bases in Serbia, with automotive production and R&D facilities in Poland and battery factories in Hungary[reference:58].

4. Biotech: The $110 Billion Licensing Boom

Perhaps the most dramatic story in the “New New Three” is biotech. In the first half of 2026, Chinese innovative drug out-licensing (license-out) deals reached approximately $110 billion—already 80% of 2025’s full-year total[reference:59][reference:60]. The number of deals: 81[reference:61].

💊 Biotech by the Numbers (H1 2026)

$110B — Out-licensing deal value[reference:62]
81 — Number of deals[reference:63]
80% — Of 2025 full-year total reached in H1[reference:64]
8 of 10 — Global top 10 biopharma deals involved Chinese companies[reference:65][reference:66]
38 — Class 1 innovative drugs approved by NMPA[reference:67]
11 — “Global first” new-target drugs (all自主研发)[reference:68]
30% — China’s share of global drug R&D pipeline[reference:69]
20+ — Countries receiving licensees (US, UK, France, Italy, etc.)[reference:70]

🏆 The Mega-Deals

The first half of 2026 saw a series of blockbuster deals:

  • January: CSPC Pharmaceutical Group and AstraZeneca signed a partnership worth up to $18.5 billion—the largest single biotech licensing deal in Chinese history[reference:71].
  • May: Hengrui Medicine and Bristol Myers Squibb signed a global strategic partnership worth up to $15.2 billion[reference:72].
  • May: Innovent Biologics and Pfizer signed a cooperation agreement worth up to $10.5 billion[reference:73].

Of the top 10 global biopharma business development (BD) transactions by value in H1 2026, 8 involved Chinese companies[reference:74][reference:75]. The deals spanned 10 therapeutic areas including oncology, metabolism, immunology, and neuroscience[reference:76]. Licensees were located in 20 countries and regions including the US, UK, France, and Italy[reference:77].

🔬 From “Selling Seedlings” to “Selling Capability”

Perhaps more important than the size of the deals is the changing nature of the deals. In the past, Chinese biotech firms would “sell seedlings”—license early-stage molecules to overseas partners at low prices[reference:78]. Now, they’re co-developing and co-commercializing, participating more deeply in global R&D and market operations[reference:79].

As one industry observer noted: “Multinational pharma companies are no longer satisfied with buying a single molecule—they’re now buying portfolios. Chinese companies are no longer just selling regional rights—they’re engaging in joint development and co-commercialization”[reference:80].

The upfront payments tell the story. In the CSPC-AstraZeneca deal, the upfront payment was $1.2 billion; in the Hengrui-BMS deal, upfront and anniversary payments totaled $950 million[reference:81]. Higher upfront payments signal that multinational buyers are willing to put more money on the table early—a sign of increasing bargaining power for Chinese assets[reference:82].

🌍 The “China Moment” in Global Pharma

A drug called “selective orexin type 2 receptor agonist”—developed to treat narcolepsy, a rare condition that causes people to fall asleep suddenly—was approved in China before the US, Japan, or Europe[reference:83][reference:84]. As a National Medical Products Administration official put it: this is the “China moment” for global innovative drugs[reference:85].

“It used to be that China waited for the world’s new drugs,” one observer wrote. “Now the world waits for China’s approvals”[reference:86].

China now accounts for approximately 30% of the global drug R&D pipeline, ranking second in the world[reference:87]. In the first half of 2026, the National Medical Products Administration approved 38 Class 1 innovative drugs, of which 11 were “global first” new-target drugs developed entirely by domestic companies[reference:88].

5. From Selling Products to Selling Capabilities

The most important story behind the “New New Three” isn’t just the numbers—it’s the fundamental change in China’s export logic[reference:89].

In the past, exports meant shipping containers filled with clothes, furniture, and appliances. The buyer paid, took delivery, and the transaction was complete[reference:90]. Today, “New New Three” exports are different in kind, not just degree.

🧠

AI: Exporting Algorithms

In Peru, Chinese algorithms optimize port loading plans at Qiankai Port[reference:91]. In Brazil, Chinese AI models help local power grids solve inspection challenges[reference:92]. In Southeast Asia, Chinese AI is applied in agriculture, healthcare, and finance[reference:93]. What’s being exported is token calls and algorithm iterations—the “product” no longer has a physical form[reference:94].

🤖

Robotics: Exporting Solutions

Robot exports aren’t just equipment—they’re complete intelligent control systems and maintenance solutions[reference:95]. A cloud-based robot from China is sorting packages in a UK warehouse; the same Chinese solution provider stands behind factories in Brazil and Malaysia[reference:96]. Technical teams can complete trajectory simulations remotely—the equipment can travel without the team[reference:97].

💊

Biotech: Exporting Capability

Multinational pharma companies are no longer buying a single molecule—they’re buying target discovery, clinical translation, and global registration capability[reference:98]. Chinese companies are engaging in joint development and co-commercialization, not just regional licensing[reference:99].

As one commentary put it: “China is no longer just exporting goods—it’s exporting the ability to solve problems“[reference:100].

This shift didn’t happen overnight. It’s the result of decades of investment in R&D. Over the past five years, China’s R&D spending has grown at an average annual rate of 10%, with total investment ranking second globally[reference:101]. “The ‘New New Three’ look like a rapid concept switch, but they’re actually the concentrated release of China’s long-term accumulation reaching a critical point”[reference:102].

Customs officials have noted that this evolution is changing how trade is measured. Traditional trade statistics struggle to capture the value of token calls, algorithm iterations, and intellectual property transfers[reference:103].

6. Practical Guide: Verifying Chinese Tech & Biotech Partners

With Chinese AI, robotics, and biotech companies becoming critical partners in the global innovation economy, due diligence has never been more important. Whether you’re licensing AI technology, sourcing robotics solutions, or entering a biotech partnership, here’s a practical checklist:

  • Start with the official company record. Every legitimate Chinese company has a registered profile in the National Enterprise Credit Information Publicity System. The Official Enterprise Credit Report provides verified information on legal status, registered capital, shareholders, directors, and operational history. This is essential for any high-value technology or biotech partnership.
  • Go deeper with professional due diligence. For strategic technology or biotech partnerships, you’ll need comprehensive risk assessment. A Professional Enterprise Credit Report covers 11 dimensions including financial health, legal risks, supply chain information, and executive backgrounds.
  • Verify intellectual property. In AI, robotics, and biotech, IP is the core asset. Patent and trademark verification is essential to confirm that your partner actually owns the technology they’re licensing or selling. This is especially critical for biotech licensing deals where patent portfolios can be worth billions.
  • Check executive backgrounds. Key personnel—directors, supervisors, and senior management—can be vetted through Executive Risk Reports that reveal external investments, other positions held, and any adverse records.
  • Don’t overlook document authentication. Licensing agreements, technology transfer contracts, and other legal documents from China typically require apostille or legalization to be recognized in your home jurisdiction.

ChinaBizInsight specializes in helping overseas clients know their Chinese partners—providing everything from basic company searches to comprehensive due diligence reports and document authentication services. In an era where Chinese AI, robotics, and biotech companies are becoming essential to global innovation, having reliable information about your partners isn’t just good practice—it’s essential for protecting your investment and your IP.

Final Thoughts

The “New New Three”—AI, robotics, and biotech—have emerged as China’s next export powerhouses. In the first half of 2026, AI-related exports surpassed $480 billion, industrial robot exports reached 6.29 billion yuan across 141 countries, and biotech licensing deals hit $110 billion with Chinese companies claiming 8 of the top 10 global biopharma transactions.

But the real story isn’t just the numbers—it’s the fundamental shift in China’s export logic. From selling physical products to selling algorithms, solutions, and innovation capabilities, China’s exports have moved from “cost-driven” to “innovation-driven”[reference:104]. The “New New Three” represent a qualitative leap in China’s global economic role.

For overseas businesses, this presents unprecedented opportunities—but also new due diligence challenges. The companies you partner with today in AI, robotics, or biotech are likely to be some of the most innovative and valuable companies in their fields. But they also require more thorough verification of their IP, their executives, and their legal status.

Know your Chinese partners. The “New New Three” are here—and they’re changing the world.

C
ChinaBizInsight Editorial Team
Know Your Chinese Partners — cnbizinsight.com

📚 References

  1. General Administration of Customs of China, H1 2026 import/export press conference, July 14, 2026
  2. National Medical Products Administration (NMPA), H1 2026 innovative drug approval data, July 2026
  3. OpenRouter, weekly AI model token call data, various dates 2026
  4. Hugging Face, Spring 2026 Global Open-Source AI Ecosystem Report
  5. China.com.cn, “好评中国|’新新三样’新在何处②发展动力之新,” August 14, 2026
  6. People’s Daily / gov.cn, “开局观察” series, July 2026
  7. Xinhua News Agency / CCTV, various reports on “New New Three,” July–August 2026
  8. Northeast Securities, “China Manufacturing Export Panorama,” August 2026
  9. Various industry sources: CSPC-AstraZeneca, Hengrui-BMS, Innovent-Pfizer deal announcements, January–May 2026

Your strategic bridge to transparent business in China.

Native Expertise
Direct Access
Official Sources
VIEW SAMPLES CONSULT EXPERT

Leave a Comment

Your email address will not be published. Required fields are marked *

Scroll to Top