Alibaba

ByteDance

CATL

Huawei

Meituan

Tencent

Xiaomi

Most Followed

Market News

Listings (IPO)

Earnings Reports

Venture Roundup

Big Tech

Automotives

Retail and Commerce

Collections

News

Features

Insights

Press Releases

Anta

DeepSeek

Geely

Li Auto

Mixue Bingcheng

Pop Mart

Going Global

Latest

Collaborate

Earnings

Auto

Retail & Commerce

… phase, reducing the abruptness of waking from deep sleep. When a user gets up at night, the lamp turns on dimly and switches off again when the user returns to bed.
In the first 48 hours after launching on Kickstarter, Sleepal raised more than USD 200,000, outperforming 90% of products on the platform, according to Fan.
Fan said Sleepal’s short-term business model will combine hardware sales with software subscriptions. In the long term, the company plans to expand from sleep detection into home-based health AI solutions. The multimodal sensing technologies validated in the bedroom could, for example, be applied in bathrooms for fall detection using millimeter-wave radar, or in dining areas for diet detection.
Fan ultimately wants continuous vital sign data to support chronic disease risk screening and prevention, turning the product into an entry point for at-home health management.
The bedside lamp is only its first step.
KrASIA features translated and adapted content that was originally published by 36Kr. This <a href="https://36kr.com/p/3844006775360002">article</a> was written by Qiu Xiaofen for 36Kr.

Can a bedside lamp track sleep? A Xiaomi veteran thinks so

East Asia

China

A popular singer who was hired as the face of a <a href="/company/byd">BYD</a> brand has become a shareholder of the company, in a case that is garnering attention as a new advertising method in China's increasingly cutthroat auto market.
"I want to bring China's original music and technology, and the Chinese power that symbolizes this era, to a wider global stage," Wang Leehom said at launch event in Guangdong for BYD's high-end Denza brand last month.
On the same day, Denza announced it had appointed Wang, who debuted in 1995 and carries a wide fan base of both men and women, as the face of their brand.
Wang revealed that he purchased shares in BYD using the fees he received from the company.
"Not only did I gain a deeper understanding of BYD and feel that we share the same vision, but I also sincerely respect this company," Wang said when asked why he become a shareholder.
Celebrity spokespeople revealing that they acquired shares in that company is rare. Becoming a shareholder means that the company's earnings will affect Wang's own income, and that he will bear some responsibility as a spokesperson. Chinese media described it as "the most audacious form of 'investment-type advertising' among automakers."
Until recently, Chinese automotive advertising has been largely focused on vehicles' functions, but marketing using celebrities has been accelerating. At Beijing's Auto China exhibition in late April, actor Bai Jingting appeared at the launch event for an automotive brand involving telecommunications giant <a href="/company/huawei">Huawei</a>, eliciting screams from female fans.
There were 193 celebrities employed as spokespeople in China's transportation industry, which includes automobiles, last year, according to consumer data analysis company Endata. The figure increased 71% from 2024, the highest growth among all industries, surpassing apparel and internet.
"Automobile companies are shifting from promoting functionality to appealing to emotional connections through stars," Endata said.
<a href="/company/li-auto">Li Auto</a> announced last September that film star Jackson Yee would be appearing in its ads, the first time the company has used a celebrity as a spokesperson since its 2015 founding.
<a href="/company/xpeng">Xpeng</a> featured cellist Ouyang Nana, 25, at a launch event for a new <a href="/industry/ev">electric vehicle</a> in May 2025 in a bid to appeal to younger consumers.
Li Xiang, CEO of Li Auto, and He Xiaopeng, CEO of Xpeng, are both young founders that moved to the automotive sector from IT companies. They both symbolize the shift from gas-powered vehicles to new energy vehicles. And they both included their name in their companies, become the de facto faces of the brands.
But by 2025, the share of new energy vehicles in the Chinese market had risen to 48%, leading to a sense of saturation. As companies compete on features such as range, charging speed and high-performance driver assistance functions for EVs and plug-in hybrids, features and designs have become homogenized.
China's automakers are now urgently trying to cultivate new customer segments by redefining their brand characteristics and stories through celebrities.
The US-born Wang's emergence as a BYD shareholder comes as the company faces geopolitical pressure. Earlier this month, the US Department of Defense <a title="" href="https://asia.nikkei.com/politics/international-relations/us-china-tensions/pentagon-blacklists-alibaba-byd-and-baidu-over-alleged-military-ties" target="_blank" rel="noopener">listed the automaker</a> among companies it says have ties to the Chinese military. BYD said the move "lacks factual basis."
This article first appeared on <a href="https://asia.nikkei.com/spotlight/electric-cars-in-china/byd-brand-ambassador-wang-leehom-uses-fee-to-buy-shares">Nikkei Asia</a>. It has been republished here as part of 36Kr’s ongoing partnership with <a href="https://www.rwegrgdf5.eu.org/nikkei-teams-up-with-36kr-in-asia-tech-news-coverage/">Nikkei</a>.

BYD brand ambassador Wang Leehom uses fee to buy shares

Transportation

Electric Vehicles

Mobility

On June 9, Saidou Technology, a company backed by Seres, launched its new brand AIVA in Beijing. It also announced a partnership with Volcano Engine, a <a href="/company/bytedance">ByteDance</a> subsidiary, to jointly define and design what it calls an <a href="/industry/ai">artificial intelligence</a>-native vehicle experience, meaning one built around AI capabilities from the start.
Volcano Engine will provide AIVA with technical services, including the Doubao large language model and intelligent cockpit capabilities, according to the company. AIVA said the partnership is intended to improve in-vehicle interaction.
At the launch event, AIVA brand president and product manager Li Bo unveiled the brand’s name and logo. AIVA stands for “artificial intelligence voyage ahead,” meaning AI will accompany users as they move forward, according to the brand.
AIVA also introduced its vehicle development approach, which it calls “AI before the car.” The idea is to define each product around AI from the earliest stage of vehicle development.
Li said AIVA’s “AI-defined vehicles” will be developed by first building an AI-native organization, then using that foundation to create AI-native vehicle products.
Within that organization, AI is intended to be applied across three layers: perception, decision-making, and execution. AIVA said this structure will create a feedback loop that allows the organization to keep learning and improving.
At the product level, AIVA said AI can be used for data analysis, trend identification, needs assessment, and reasoning. Product managers could then use those insights to select development directions with clearer value. The company described vehicles developed this way as products shaped by AI’s understanding of users, with emotional warmth and adaptability across different scenarios.
AIVA also unveiled its first concept car, the Origin Concept, as an expression of the brand’s philosophy. The company said the concept starts from a large model and uses biomechanics to reinterpret the vehicle’s overall form, giving AI a physical expression that can be seen, heard, perceived, and responded to.
<figure id="attachment_162675" aria-describedby="caption-attachment-162675" style="width: 1920px" class="wp-caption alignnone"><img loading="lazy" decoding="async" class="size-full wp-image-162675" src="https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept.jpg" alt="" width="1920" height="1080" srcset="https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept.jpg 1920w, https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept-720x405.jpg 720w, https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept-1024x576.jpg 1024w, https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept-768x432.jpg 768w, https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept-1536x864.jpg 1536w, https://console.rwegrgdf5.eu.org/wp-content/uploads/2026/06/AIVA-Origin-Concept-480x270.jpg 480w" sizes="(max-width: 1920px) 100vw, 1920px" /><figcaption id="caption-attachment-162675" class="wp-caption-text">Photo source: Saidou Technology.</figcaption></figure>
AIVA’s first production vehicle, the ME7, will be unveiled in 2026, according to the company. Its full lineup will target the mainstream market above RMB 200,000 (USD 29,546).
The brand said it will maintain an efficient pace of product iteration and continue expanding its portfolio. All models will be refined around four core areas: AI interaction, emotional companionship, intelligent experience, and driving safety.
“2026 is being seen as the first year of physical AI. As a carrier equipped with sensors and actuators, the car is a core scenario for the implementation of physical AI,” said Yang Liwei, vice president of Volcano Engine. Yang added that he sees AI evolving from turn-based chatbots toward agentic AI and physical AI.
According to company materials, AIVA is not operated by a single corporate entity. Instead, it is described as an independently operated brand coordinated across several industrial partners:
<ul>
<li aria-level="1">AIVA is responsible for branding, products, and business development.</li>
<li aria-level="1">Seres provides support in vehicle manufacturing, supply chain management, quality systems, and engineering.</li>
<li aria-level="1"><a href="/company/catl">Contemporary Amperex Technology</a> (CATL) provides energy solutions covering battery technology, battery safety, and power battery charging and swapping systems.</li>
<li aria-level="1">A state-owned capital platform provides industrial resources, while Volcano Engine provides the technical foundation for AIVA’s AI capabilities and supports the optimization of its in-vehicle interaction experience.</li>
</ul>
Public information shows that Saidou was formed in late May through the renaming and restructuring of Landian Technology, which was previously under Seres. Chongqing state-owned asset managers together hold 34.5% of Saidou, making them its largest shareholder group. Seres holds 32.96%, ranking second, while CATL holds 9.89%, ranking fourth.
AIVA’s brand team also has ties to Seres. Chairman Zhang Zhengyuan serves as an employee director and assistant vice president at Seres, while Li was previously a vice president of the automotive group.
This <a href="https://xueqiu.com/2854806290/394026687">article</a> was adapted based on a feature originally written by SY and published on IPO Zaozhidao. KrASIA is authorized to translate, adapt, and publish its contents.
Note: RMB figures are converted to USD at rates of RMB 6.77 = USD 1 based on estimates as of June 15, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates.

Seres-backed AIVA debuts, with first car model set for unveiling this year

Seres

Luo Fuli, head of <a href="/company/xiaomi">Xiaomi’s</a> MiMo large model team, said Anthropic’s Claude Fable 5 remains an interim-stage product, even as the model has drawn attention for its coding and agent capabilities.
Luo made the comments while discussing Fable 5 at the eighth Beijing Academy of Artificial Intelligence (BAAI) Conference, held on June 12.
Anthropic launched Claude Fable 5 on June 9, describing it as its most capable generally available <a href="/industry/ai">artificial intelligence</a> model and saying it was built for demanding reasoning and long-horizon agentic work. The company said the model showed state-of-the-art performance on nearly all tested benchmarks, with particular strength in software engineering, knowledge work, vision, scientific research, and other complex tasks.
In one <a href="https://www.anthropic.com/news/claude-fable-5-mythos-5">example</a> cited by Anthropic, Stripe reported that Fable 5 completed a codebase-wide migration in a 50 million-line Ruby codebase in one day, a project that would otherwise have taken a full team more than two months by hand.
Luo said Fable 5 marks a substantial improvement in coding and agent capabilities, but she argued that it is still best understood as the result of continued scaling across three dimensions:
<ul>
<li aria-level="1">The first is parameter scale. Luo said she speculates that Fable 5’s parameter count may be several times larger than that of today’s strongest open-source models, reflecting further scale-up during pretraining.</li>
<li aria-level="1">The second is computing power. She said the model appears to have required significant compute for test-time scaling and reinforcement learning.</li>
<li aria-level="1">The third is data. As AI moves from the chat era into the agent era, Luo said training data is also changing. Model training is expanding beyond internet text to synthetic data generated jointly by humans and agents, pushing data scale into a new order of magnitude. In the past, she said, the full corpus of internet text data may have involved 40–80 trillion unique tokens. Today, that scale has entered a new phase.</li>
</ul>
“Fable 5 is the product of a very natural and outward extension of large models across three dimensions: parameter scale during pretraining, data, including agent-generated synthetic data, and the combination of post-training and reinforcement learning. It is an interim model,” Luo said.
As foundation models and agent technologies advance, self-evolution has become a recurring topic in the industry. Anthropic recently published a post titled <a href="https://www.anthropic.com/institute/recursive-self-improvement">“When AI builds itself,”</a> outlining recursive self-improvement, or the possibility that AI systems could eventually automate the design and development of next-generation AI systems. Anthropic said that outcome is not here and is not inevitable, but could arrive sooner than institutions are prepared for.
Asked how she views model self-improvement, Luo said today’s leading models can already solve some abstract problems at the execution layer. By contrast, the previous generation of leading models could execute well only when human instructions were clear.
She pointed to how models are being used in research workflows. A complete research process involves proposing hypotheses, designing experiments, carrying them out, validating results, communicating with peers, and revising one’s views. Today’s large models, Luo said, can already design reasonable validation metrics, verify the accuracy of their own execution results, and plan the overall experimental process.
The main gap between models and top researchers, she said, lies in their ability to propose hypotheses or raise questions worth testing. That ability amounts to research judgment, including knowing when to stop pursuing unproductive directions. Luo said stronger models, combined with better recursive improvement systems, are gradually pushing that gap toward the frontier.
If AI’s capacity for self-improvement continues to strengthen, the next question is how it will reshape the world. Will it first reconstruct the digital world before entering the physical world, or will it regain an understanding of foundation models through the physical world?
Luo said language models and world models are advancing in parallel, but language models are moving faster for now. She said their current capabilities are better suited to recreating the conditions in which intelligence emerges from agent environments.
“We use a set of agent systems that can drive models to reach a higher ceiling, then layer models on top of them and let them freely explore in an environment. This allows us to design a more precise incentive system to drive models’ self-improvement,” she said. This path, she added, is already unfolding in the digital world.
In Luo’s view, the world model layer has yet to produce a highly efficient vision model. The first step should be an efficient generator architecture, followed by a scaffolded agent system capable of handling complex real-world tasks. The full paradigm then needs to be scaled. Language models, she said, are likely to clarify this path first.
This <a href="https://xueqiu.com/2854806290/394616477">article</a> was adapted based on a feature originally written by SY and published on IPO Zaozhidao. KrASIA is authorized to translate, adapt, and publish its contents.

Xiaomi’s Luo Fuli says Claude Fable 5 is an interim step in AI’s evolution

Artificial Intelligence

Deep Tech

Foundation and Large Models

Japan, South Korea, and other big Asian economies have been dealt a blow by the Iran war's disruption of naphtha imports—all except China, which has weathered the crisis thanks to its diversified supply networks.
"The recent blockade of the Strait of Hormuz has emerged as a major crisis that exposes the Asian petrochemical industry’s structural vulnerabilities," said Baek Jong-hoon, acting chairman of the Korea Chemical Industry Association and CEO of Kumho Petrochemical, at an industry event in Fukuoka, Japan, last month.
The de facto closure of Hormuz, through which about 20% of global crude oil exports transit, has hit Asia the hardest. Over 80% of crude oil passing through the maritime chokepoint was bound for Asian markets, according to the US Energy Information Administration.
Many Asian countries, including Japan and South Korea, rely heavily on imported naphtha, produced when crude oil is refined, to make chemicals for plastics, inks and other products.
Japan has cut production at ethylene plants that use naphtha, while some plants in South Korea and Vietnam have been shut down. All these countries are trying to maintain supply chains by procuring naphtha from outside the Middle East.
But China, which boasts the world's largest chemical production capacity, is feeling less of a pinch. Its crude oil sources are less concentrated in the Middle East and include Russia. Chinese chemical companies also use feedstocks other than naphtha, including ethane derived from natural gas and coal.
China can continue to supply chemicals even if Middle Eastern naphtha supplies are cut off.
Prices of naphtha from outside the Middle East, which at one point doubled compared with before the conflict, have eased somewhat but remain about 30% higher. If this situation drags on, Chinese chemicals, which already have a price advantage, could dominate the market.
"The competitiveness of Asia's petrochemical industry, including Japan, will further decline," said a senior executive at a major Japanese chemical company.
Asia's petrochemical industry has long suffered from a glut of cheap imports from China. Baek said the current Middle East crisis is a double blow to the industry, which was already struggling from declining profitability.
South Korea, which has higher production capacity than Japan, is promoting the consolidation of ethylene plants under government leadership.
In Thailand, state-owned oil company PTT and Siam Cement Group, the country's largest materials maker, said in April that they will explore merging their basic chemicals and commodity-grade plastic businesses.
"Even before recent disruption, the sector had already been shifting toward leaner capacity, higher utilization rates, and greater emphasis on specialty and high-value products," said Toasaporn Boonyapipat, chairman of a Thai petrochemical industry association. "Today, that transition is moving faster and with greater urgency."
In response to disruption caused by the Middle East conflict, the Japanese government has spearheaded POWERR Asia, a proposal to work with Asian countries on resource supply chains.
"In an era defined by geopolitical uncertainty, overcapacity, and the urgent call for climate action, collaboration is no longer a choice, it is essential," said Tsao Mihn, chairman of a Taiwan industry association and chairman of Formosa Petrochemical. "Only by working together can we turn these challenges into opportunities."
Japanese chemical companies have restructured to bring their capacity in line with shrinking domestic demand. But Koshiro Kudo, chairman of the Japan Petrochemical Industry Association and president of Asahi Kasei, said that "with unexpected changes possible, it is also necessary to have a buffer and respond flexibly."
Kudo argues for a framework in which Asian economies facing similar challenges complement each other.
"Even just exploring risk diversification through exchanges with geographically close South Korea would be very meaningful," Kudo said.
This article first appeared on <a href="https://asia.nikkei.com/business/markets/commodities/what-naphtha-shortage-china-extends-petrochemicals-lead-during-iran-war">Nikkei Asia</a>. It has been republished here as part of 36Kr’s ongoing partnership with <a href="https://www.rwegrgdf5.eu.org/nikkei-teams-up-with-36kr-in-asia-tech-news-coverage/">Nikkei</a>.

What naphtha shortage? China extends petrochemicals lead amid US-Iran conflict

Japan

… for the California assembly plant and a second manufacturing facility were delayed in the second half of 2025 by employment disputes. He said Windrose was in the process of selecting a new site in California for truck production.
Former staffers said the original California site was little more than a warehouse, too small for production, and that a dispute with the landlord forced the company out in June 2025, months after signing the lease. They said the disagreement stemmed from unpaid rent.
Han, however, blamed a consultant representing Windrose in negotiations, saying his conduct alarmed the landlord and led to the company forfeiting USD 200,000 in prepaid rent. The landlord declined to comment.
Currently, the company's electric trucks are fully assembled in factories contracted by Windrose in the Chinese city of Suzhou, using components from suppliers including battery maker CALB, Zhenghe Motor, and Hande Axle, according to Han.
Han told Nikkei Asia in 2025 that he had secured 200 committed orders globally for made-in-China trucks, which he said were already authorized by governments to be on the roads in Asia, Europe, and North and South America. Han said the company has since delivered 20 trucks to customers.
Former employees said Windrose shipped just six trucks to the US from China during 2025, four of which were not road-certified. Customers were not allowed to test drive the trucks or haul freight, they said.
"We had to pull a trailer around with blocks in the back to show the weight, but a customer couldn't drive it. It had to be a Windrose employee," said a former worker, adding that this impacted potential sales.
Former senior employees told Nikkei Asia they pressed Han several times for financial documents so that they would be able to put together a business strategy and budget. "We were trying to get transparency to that, and the root cause of the problem, and adjust our plans," one former employee said. Han refused each time, he said.
Han said he shared some financial statements with certain employees, but added, "It is not my duty to show every number that every employee wants at any point in time."
Pursuing international growth, Windrose made Antwerp, Belgium, its global headquarters in 2024. It sponsored the local basketball team, now the Windrose Giants, and planned to build an assembly facility at the Port of Antwerp-Bruges.
Through a partnership announced last month, Han said the company was also prepared to do final assembly at Belgian logistics company Aertssen's American factories in Tacoma, Washington, and Rincon, Georgia. Han said he did not know the capacity of the facilities. Aertssen did not respond to multiple requests for comment.
In April, Windrose said it delivered its first electric truck to US customers, Texas-based logistics company Allogic and its charging partner Greenspace, through its official distributor and aftermarket partner, Xos. Xos did not respond to a request for comment.
Allogic and Greenspace told Nikkei Asia in June they were expecting a second Windrose truck in the coming months and had made no commitment to purchase more, as they looked to raise funds.
Han said his company aimed to ship 20 made-in-China trucks to the US in the third quarter of this year, despite steep tariff rates of at least 50%.
"Last year was like a great reset for what we do," Han said. "Everything is happening this year."
This article first appeared on <a href="https://asia.nikkei.com/business/automobiles/electric-vehicles/chinese-entrepreneur-s-e-truck-startup-windrose-faces-unpaid-wage-claims">Nikkei Asia</a>. It has been republished here as part of 36Kr’s ongoing partnership with <a href="https://www.rwegrgdf5.eu.org/nikkei-teams-up-with-36kr-in-asia-tech-news-coverage/">Nikkei</a>.

Chinese entrepreneur's e-truck startup Windrose faces unpaid wage claims

… a consumer analyst at investment bank CITIC CLSA. "More sports categories are emerging, which means multi-brand is a better strategy to address consumer demands."
Anta's global ambitions are just as strong. As early as 2005, the company's founder, Ding Shizhong, an entrepreneur who hails from Jinjiang, a shoe manufacturing town in Fujian, had rejected the suggestion that it would be only a "Nike" for China, setting his sights on markets worldwide.
"What Anta is doing—using acquisitions and partnerships to go global—is increasingly the playbook for ambitious Chinese consumer brands," IMG's Zhu said.
Last month, fast fashion e-commerce giant <a href="/company/shein">Shein</a> said it would acquire Everlane, a US retailer best known for marketing its apparel as ethically sourced and sustainable. Earlier this year, Centurium Capital, the private equity firm backing China's budget coffee chain Luckin Coffee, bought US roaster Blue Bottle.
Anta's Beverley Hills venture shows it is also trying to go up-market and change perceptions of the Chinese brand, while retaining its reputation for low entry price points and value for money. The latest Anta basketball shoe collection KAI 3—a collaboration with Kyrie Irving—sells for USD 135 a pair, compared to a pair in Nike's LeBron collection, named after NBA legend LeBron James, costing USD 210–235.
But the ultimate goal is to use the knowledge and experience gained from international acquisitions to expand recognition and sales of the Anta brand well beyond Beverly Hills and basketball fans.
"[The aim] is to share technology, product know-how, and market insights across the portfolio," said Zhu, "So the core Anta brand becomes more globally competitive."
This article first appeared on <a href="https://asia.nikkei.com/business/business-asia/how-china-s-anta-joined-nike-and-adidas-on-the-world-sportswear-podium2">Nikkei Asia</a>. It has been republished here as part of 36Kr’s ongoing partnership with <a href="https://www.rwegrgdf5.eu.org/nikkei-teams-up-with-36kr-in-asia-tech-news-coverage/">Nikkei</a>.
Note: EUR, RMB figures are converted to USD at rates of EUR 0.87 = USD 1 and RMB 6.79 = USD 1 based on estimates as of June 8, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates.

How China's Anta joined Nike and Adidas on the world sportswear podium

Retail

Consumer Economy

ByteDance’s <a href="/industry/ai">artificial intelligence</a> drug discovery unit has begun the process of spinning off and raising independent financing, according to people familiar with the matter.
<a href="/company/bytedance">ByteDance</a> will retain a controlling stake in the new company after the spinoff, the people said. The AI drug discovery unit’s team, algorithms, technology platform, and existing pipeline assets will be transferred into the new entity. The business will also continue to receive computing power support from Volcano Engine.
Founded in 2021 and led by Kai Liu, the unit has about 50 core members, sources told 36Kr. The team includes AI-for-science (AI4S) talent, algorithm specialists, and pharmaceutical experts. Since its launch, it has worked across foundation model research, platform development, and commercialization.
ByteDance’s team previously responsible for protein structure prediction models has also been folded into the AI drug discovery team led by Liu. The related algorithm and model teams have completed their integration and will continue advancing foundation model research in the field. A small number of employees have left.
The progress of ByteDance’s AI drug discovery unit has given the company a basis to pursue the spinoff financing.
ByteDance has produced several technical outputs in AI drug discovery. In 2025, its AI4S team released the molecular structure prediction models Protenix and Seedfold. In 2026, it released Protenix-v1 and Protenix-v2, building open-source structure prediction capabilities for biomolecular systems such as proteins and ligands.
In protein design and prediction, the team launched PXDesign and other tools for designing protein binders.
ByteDance has also launched Anew Labs, an AI-powered platform focused on real-world drug development.
According to Anew Labs’ website, the team has released research including AnewSampling, AnewOmni, AnewFEP, AnewSynth, and scNext. These projects cover areas such as protein-ligand dynamic structure prediction, all-atom molecule generation, free energy calculation, synthetic feasibility prediction, and virtual cells. The platform has also launched early-stage drug pipelines.
In April, Anew Labs disclosed its IL-17 small-molecule program for the first time at the annual meeting of the American Association of Immunologists. According to the disclosure, it was the first reported case globally of using small molecules to block three IL-17 family dimers, or paired protein complexes. IL-17 is an important pathway in autoimmune diseases such as psoriasis and ankylosing spondylitis. Antibody drugs have already clinically validated the simultaneous inhibition of two key inflammatory cytokines.
As ByteDance’s AI drug discovery work has matured, the company is trying to move the effort from scientific research toward industrial application. It has integrated related internal teams and is preparing to test commercialization.
Commercializing AI4S remains difficult.
AI4S businesses have long validation cycles and complex development processes. Drug discovery, for example, spans model R&amp;D, wet-lab experiments, and clinical validation. That requires specialized talent and an organizational model that differs from ByteDance’s internet businesses.
People familiar with the matter said the spinoff is intended to establish an independent structure better suited to the business. ByteDance hopes the adjustment will help it attract top talent and further integrate foundation model capabilities, algorithms, and pharmaceutical expertise.
The broader drug development sector is also under pressure to improve efficiency.
Over the past two decades, global pharmaceutical companies have continued to increase R&amp;D spending. IQVIA, a healthcare data and clinical research services provider, expects global spending on medicines to reach about USD 2.3 trillion by 2028.
The market is large, but the core problems of new drug development, including high costs, long cycles, and high failure rates, have not fundamentally changed. This has created demand for AI tools that could improve parts of the process.
AI4S research is also accelerating, with models becoming more capable of addressing complex scientific problems.
AlphaFold, Google DeepMind’s protein structure prediction model series, offers one example. The first generation demonstrated feasibility. AlphaFold 2 expanded protein structure prediction at scale, with predicted structures for about 200 million proteins. AlphaFold 3 then extended beyond single-protein prediction to model more complex interacting systems. The progression shows how AI is becoming increasingly relevant to drug design.
If protein structure prediction remains a fundamental research question, multimodal molecular generation models that have emerged in recent years are addressing a more direct pharmaceutical challenge: drug design. That suggests AI drug discovery is gradually moving from research toward industrial application.
ByteDance has been investing in AI4S for years. Around 2020, it began exploring areas such as AI drug discovery, molecular simulation, and computational biology. Since then, it has built teams covering first-principles calculations, quantum chemistry, molecular dynamics, materials simulation, and molecule generation for energy and drug applications.
After ByteDance established Seed, its foundation model research team, AI4S became part of the company’s frontier technology strategy.
A person close to the spinoff said this is ByteDance’s first attempt to commercialize AI4S. The person added that the company is placing significant internal emphasis on the effort, in the hope that more independent decision-making can help it find a viable path in the field.
KrASIA features translated and adapted content that was originally published by 36Kr. This <a href="https://36kr.com/p/3846956646124036">article</a> was written by Zhou Xinyu for 36Kr.

ByteDance spins out drug discovery unit to test AI4S commercialization

Biotechnology

… latest SUVs look similar in price and configuration, but each pushes one part of the experience toward the upper end.
Li Auto’s new Li L9 collected 10,000 locked orders within two weeks of launch, while the new Aito M9 passed 20,000 locked orders within 24 hours. Nio’s ES9 currently shows a delivery wait of 16–18 weeks.
Early order numbers matter, but user structure and brand trajectory matter more. 36Kr previously reported that returning owners accounted for 70% of Li L9 Livis orders in the first 72 hours. Nio salespeople said the ES9’s repeat purchase rate is also high, with nearly half of orders coming from existing owners.
The two brands’ existing customer bases differ. Li Auto’s users prioritize comfort but are also conscious of value for money. Some salespeople said not all existing owners can afford RMB 500,000 models, and many are waiting for cheaper Li L7 and L8 Livis versions. Nio’s customer base is more premium. The ES8, priced above RMB 500,000, exceeded 10,000 annual sales in 2018, two years before Li Auto launched its first model in the RMB 300,000–350,000 (USD 44,213.2–51,582.1) range.
Store visits showed that 20–30% of new ES9 orders were corporate purchases. Even with zero down payment and five-year installment plans with two years interest-free, full-payment purchases remain common. Li Auto, by contrast, has more installment buyers.
Aito resembles Nio, but its premium recognition appears broader. Its influence reaches beyond first- and second-tier cities into lower-tier markets. One Aito salesperson said county-level business owners may prefer Aito because, in business settings, people recognize it in a way they do not recognize Li Auto or Nio.
Nio’s sales are closely linked to its battery swap network, giving it a high-end user profile but also a concentrated market distribution in first- and second-tier cities. Traditional flagship SUVs such as the BMW X5 have broader and more stable influence, covering more than 100 cities even as monthly sales fluctuate.
To take root in the high-end market, Chinese NEV brands need symbols that extend beyond core users and early adopters. For Li Auto, that symbol remains family comfort. For Nio, it is design, service, and owner identity. And for Aito, it is Huawei-backed intelligence and trust.
KrASIA features translated and adapted content that was originally published by 36Kr. This <a href="https://36kr.com/p/3839803361741313">article</a> was written by Xu Caiyu for 36Kr.
Note: RMB figures are converted to USD at rates of RMB 6.79 = USD 1 based on estimates as of June 10, 2026, unless otherwise stated. USD conversions are presented for ease of reference and may not fully match prevailing exchange rates.

China’s electric SUV makers chase high-end buyers

Aito