CHONGQING, China — Alibaba on Wednesday announced a new artificial intelligence chip that is three times more powerful than its predecessor, adding to China’s domestic AI chip options as Nvidia struggles to get its products into the country.

The Zhenwu M890 delivers three times the performance of the current Zhenwu 810E, Alibaba said, adding that the new processor has 144 GB GPU memory and interchip bandwidth of 800 GB per second.

The e-commerce and technology giant said it had already delivered 560,000 Zhenwu units to more than 400 customers across 20 industries.

The new chip could make Alibaba and its chip subsidiary T-Head more competitive in China’s growing domestic AI processor market, which includes competitors like Huawei and Cambricon.

“Alibaba designed AI chips are making headway with external customers and are becoming one of the more popular platforms among Chinese domestic AI hardware chips,” said Myron Xie, an analyst at SemiAnalysis with a focus on AI accelerators.

However, he noted that the advertised memory capacity and bandwidth figures are still lagging behind those of major Western chip companies. Alibaba has yet to release other important metrics such as compute performance, Xie added.

Chinese AI developers have long been restricted from buying cutting-edge processors from companies such as Nvidia due to American export restrictions.

Beijing has also tightened scrutiny on domestic companies’ use of foreign AI chips, including Nvidia’s H200 chip, despite Washington recently clearing their sale in China.

“Given that Nvidia remains out of China … [it is] unlikely that Nvidia will be a long-term supplier into all of China,” said Leonid Mironov, portfolio manager at Gavekal.

Alibaba’s chip announcement also suggests investors should not discount Alibaba and Tencent, Mironov said. The two companies are the largest holdings in his newly launched China fund.

“Alibaba is also keeping pace and doing really well with T-head,” he said.

The Zhenwu will give local tech companies another option, although questions remain over how much manufacturing capacity Alibaba can continue to secure at local semiconductor foundries such as Semiconductor Manufacturing International Corporation, according to Counterpoint Research Associate Director Brady Wang.

“M890 is a small but real contribution to China’s AI self-sufficiency … On raw silicon power, M890 is not a true competitor to H200. But it does not need to be. In the China market, it is a believable replacement for H200,” Wang added.

Alibaba’s latest AI processor could also help support the computing demands of its Qwen large language models. The company is widely considered a “full-stack AI company,” due to its business spanning across hardware, computing, AI models, tools and applications.

Alibaba also revealed on Wednesday that its next-generation AI model, Qwen3.7-Max, would soon be released.

In early April, Alibaba and China Telecom said they were launching a data center in southern China powered by the e-commerce giant’s own chips.

— CNBC’s Arjun Kharpal contributed to this report.

BEIJING — Alibaba Cloud is investing in a new type of artificial intelligence designed to better replicate the real world using a different approach from chatbots such as OpenAI’s ChatGPT.

The shift recognizes the limits of “large language models” trained primarily on text. Instead, developers are starting to focus more on “world models” built on videos and real-life physical scenarios.

To jump on the trend, Alibaba led a 2 billion yuan ($290 million) investment in ShengShu, the startup behind the AI video generation tool Vidu, the company announced Friday. TAL Education and Baidu Ventures also participated in the series B funding round.

The investment comes about two months after ShengShu raised 600 million yuan from Qiming Venture Partners and other backers. The startup declined to disclose its valuation.

ShengShu said the latest funding will support the development of a “general world model” that uses AI to bridge two currently separate domains: the digital world of games and AI-generated video, and the physical world of autonomous driving and robots.

“ShengShu believes that a general world model, built on multimodal data such as vision, audio, and touch, more naturally captures how the physical world works than large language models,” the three-year-old startup said in a statement.

“We aim to connect perception and action,” Zhu Jun, founder of ShengShu, added in a statement, allowing AI systems to better model and predict real-world behavior consistently.

ShengShu’s latest Vidu Q3 Pro model, released in January, ranks among the top 10 AI models for generating videos from text and images, according to Artificial Analysis.

The company launched Vidu globally months before OpenAI made its now-shuttered Sora tool for AI video generation widely available. Chinese short-video companies Kuaishou and ByteDance have also released similar competing AI tools for generating videos.

The Chinese tech giant and Baidu Ventures last month led a $50 million investment in Tripo AI, a platform that uses AI to quickly generate digital 3D models from photographs. Tripo said it is also moving away from techniques used by language models toward AI tools grounded in physical space and is developing its own world model.

In September, Alibaba also led a $60 million investment in PixVerse, which released an AI world model earlier this year that allows users to direct how a video unfolds while it is being generated.

Alibaba, which got its start in e-commerce, has also released free, open-source AI models for video generation and, in February, launched one for powering robots.

Shengshu said Friday it has strategic partnerships with companies developing embodied AI — systems such as humanoid robots that interact with the physical world — for use across industrial, commercial and home settings.

World models are critical for robotics because the technology needs more than LLMs to work, Kevin Kelly, co-founder of the U.S. tech magazine Wired, wrote last month on his Substack.

Ultimately, to replicate human intelligence, AI will need three things: reasoning, an understanding of the physical world and continuous learning, Kelly said. While AI for the learning category hasn’t been developed yet, LLM-powered chatbots have created the knowledge element, he said, making world models a key area requiring a breakthrough.

Alibaba Group has released its newest AI model series, featuring enhanced capabilities, as it faces intensifying competition in China’s AI space with several models launched in the past week. 

The Qwen3.5 AI model comes in an open-weight version, which allows users to download, run, fine-tune, and deploy it on their own infrastructure. Alibaba also released a “hosted version,” meaning the model can run on Alibaba’s own servers.

Both models were made available on Monday, the eve of the Chinese New Year, and come just a week after Alibaba released a new AI model designed for robots.

The company highlighted that Qwen3.5 offers improvements in performance and cost and was built with “native multimodal capabilities,” enabling the models to understand text, images and video simultaneously within one system.

Leaning into a major AI trend this year, the model also supports new coding and agentic capabilities and is compatible with open-source AI agents like those from OpenClaw, which recently surged in popularity.

AI agents are systems that can independently take actions and complete multi-step tasks on a user’s behalf with minimal supervision.

These agents and their abilities have garnered a lot of attention in recent weeks, after American AI company Anthropic released new agent tools. The potential for these agents to replace the work of software as a service companies, amongst others, has rocked markets.

Alibaba’s local competitors such as ByteDance and Zhipu AI also released upgraded models in the past week aimed at supporting more agent capabilities.

The company said that its new Qwen3.5 open-weight model comes with 397 billion parameters — variables that shape how an AI system learns and reasons. While less than its previous flagship model, the company said the latest model showed significant improvement based on self-reported benchmark evaluations.

Alibaba provided benchmark tests showing that Qwen-3.5’s performance was on par with leading models from OpenAI, Anthropic and Google DeepMind, though the comparisons were self-reported.

Meanwhile, it also released a “hosted model” called the Qwen-3.5-Plus through its cloud platform Model Studio. Alibaba said this version also demonstrated performance on par with leading competitors. CNBC could not independently verify those claims.

The new Qwen3.5 models also support 201 languages and dialects, up from the previous generation’s 82. 

Alibaba is expected to release more open-weight models during this Chinese New Year, Lin Junyang, technical lead of Alibaba Cloud’s Qwen team said in a social media post.

Following the release of Anthropic’s latest Claude AI agent tools, other American AI giants have been accelerating the development of agentic capabilities. OpenAI CEO Sam Altman said Sunday that the creator of the OpenClaw would be joining the company.

Last month, Google DeepMind head Demis Hassabis told CNBC that Chinese AI models were just “months” behind Western rivals.

BEIJING — Chinese companies are accelerating the rollout of new artificial intelligence models as competition with U.S.-based rivals OpenAI, Anthropic and Google intensifies.

Just over a year ago, China-made DeepSeek rocked global markets with the release of an AI chatbot that undercut OpenAI’s ChatGPT on usage fees and production costs, raising questions about the efficacy of U.S. tech restrictions on China.

On Tuesday, Beijing-based startup Moonshot AI revealed Kimi K2.5, which claimed to have video-generation and agentic capabilities that outperformed all three of the leading U.S. AI models. Agentic AI broadly refers to AI systems capable of carrying out tasks on behalf of people. The ultimate goal is to have sophisticated agents that work autonomously with minimal user interaction.

The update came just about three months after Moonshot released its K2 model.

Hours earlier, e-commerce giant Alibaba announced its latest generative AI model, which can create text, pictures or video based on user commands. Alibaba claimed its Qwen3-Max-Thinking outperformed major U.S. rivals on a broad benchmark test called “Humanity’s Last Exam.”

Alibaba said the new model can automatically select the best AI tool for a range of tasks and draw on past conversations as context to generate new responses more efficiently, all at little additional cost.

A week earlier on Jan. 19, Z.ai released a free version of its recently launched GLM 4.7 model. Two days later, the company restricted new subscribers from signing up for its AI coding tool after demand strained its available computing power.

Also last week, Hong Kong-traded shares of Chinese tech giant Baidu climbed to their highest level in nearly three years following the release of its latest generative AI model, Ernie 5.0. The company claimed the update outperformed Google’s Gemini-2.5-Pro, but did not compare it with Google DeepMind’s newest Gemini 3 Pro.

Demis Hassabis, the CEO of Google DeepMind, told CNBC this month that China’s AI models may be just “months” behind those developed in the U.S.

There are indications that the trend is already happening. Microsoft earlier this month noted estimates that DeepSeek usage in Africa is two to four times higher than in other regions.

Paying too much attention to AI benchmarks distracts from the tech’s real value — when it is integrated into existing gaming or entertainment ecosystems such as Tencent’s, Ivan Su, senior equity analyst at Morningstar, said in a note.

The Chinese tech giant operates the ubiquitous WeChat messaging app as well as other popular gaming and video streaming apps in China.

Tencent on Sunday announced that it will distribute 1 billion yuan ($140 million) in cash awards through its Yuanbao AI chatbot app during the Lunar New Year festival in February. The promotion mimics earlier “red envelope” campaigns that helped WeChat become one of the two dominant mobile payment apps just over a decade ago. It is customary for families to give relatives cash in small red envelopes during the holiday.

ByteDance and Baidu are also running AI-related promotions around the upcoming holiday in a bid to retain users on their AI applications.

Earlier this month, Alibaba updated its Qwen AI app to encourage users interacting with the chatbot to shop, order food and pay without leaving the app — thanks to integration with the company’s existing e-commerce platforms such as Taobao. Qwen claims more than 100 million monthly active users.

This integration could generate more revenue for Alibaba, on the assumption that the greater the number of people who use Qwen, the more they will use Taobao and shop on it, Lu said. That would help offset the costs of developing and running the underlying AI model, he said.

Chinese companies are mostly competing for user traffic rather than developing the most advanced models, he said.

BEIJING — Alibaba is helping the Chinese military to target the U.S., according to a White House memo, the Financial Times reported Friday.

The memo alleged “Alibaba provides tech support for Chinese military ‘operations’ against targets in the U.S.,” according to the FT.

The FT said it could not independently verify the claims, and did not publish a full version of the memo. It was not clear when the memo was released. The White House did not respond to requests for comment, while the FT said it stood by its reporting.

“The assertions and innuendoes in the article are completely false,” Alibaba said in a statement to CNBC on the FT report.

“We question the motivation behind the anonymous leak, which the FT admits that they cannot verify,” Alibaba said. “This malicious PR operation clearly came from a rogue voice looking to undermine President Trump’s recent trade deal with China.”

U.S. President Donald Trump and Chinese President Xi Jinping met in South Korea last month for the first time since Trump began his second term in January. The leaders agreed to a rollback of tariffs and export controls for 12 months, easing bilateral tensions that have escalated this year.

The lack of details in the FT report “does raise the question of whether some of the China hawks in the administration are trying to undercut the President’s deal with Xi Jinping,” Andy Rothman, founder of consulting firm Sinology, said Monday on CNBC’s “Squawk Box Asia.”

He pointed out that Trump had not said anything about the FT report, while noting that all the major U.S. cloud computing companies have contracts with the U.S. government.

The U.S. has ramped up efforts over the last few years to restrict China’s access to advanced semiconductors for training artificial intelligence models.

“The fact that Alibaba’s stock price dropped so rapidly in response [to the FT report] shows how much China’s AI industry is on edge over possible new sanctions,” said Kyle Chan, a Brookings fellow who focuses on China tech.

Alibaba shares had closed 3.78% lower in the U.S. on Friday following the report, but were up more than 1% in Hong Kong on Monday.

Chan pointed out the FT report comes as Alibaba’s open-source Qwen AI model is growing in popularity in Silicon Valley, increasing the threat to the pay-to-use models from U.S. AI companies OpenAI and Anthropic — while investors are increasingly worried about a possible AI bubble.

Alibaba is set to release quarterly results on Nov. 25 ahead of the U.S. market open.

—CNBC’s Eamon Javers and Elaine Yu contributed to this report.

SHANGHAI — Chinese tech giant Alibaba is already recouping its investment on artificial intelligence in the company’s e-commerce business, vice president Kaifu Zhang told reporters on Thursday.

The Chinese tech giant has bet big that AI will generate returns despite market concerns that companies are spending too much on the technology with little to show for it. Alibaba last month announced it will increase its spending on AI and cloud infrastructure, after pledging in February it would spend 380 billion yuan ($53 billion) over the next three years on the tech.

Zhang oversees e-commerce AI applications at Alibaba. Earlier in the day, he shared how the company has rolled out a range of AI tools, from making search results more personalized to improving the accuracy of virtual clothing try-ons.

The presentation comes a day after Alibaba began presales for Singles Day, China’s biggest shopping event of the year that’s akin to Black Friday.

Zhang said preliminary testing has shown consistent results from AI, including a 12% increase in returns on advertising spend.

“It’s very rare to see double-digit changes” in such tests, he said in Mandarin, translated by CNBC. Zhang predicted that thanks to AI integration, there would be a “very significant” positive impact on Alibaba’s gross merchandise volume during this year’s Singles Day shopping period, which centers on Nov. 11.

Alibaba’s China e-commerce unit remains the tech giant’s largest source of revenue, with growth of 10% year-on-year in the quarter ended June 30 to the equivalent of $19.53 billion.

Despite lackluster Chinese consumer spending in the last few years, during the Singles Day period last year, research firm Syntun estimated 20.1% year-on-year growth in sales to 1.11 trillion yuan for Alibaba’s Tmall, JD.com and PDD.

The company on a late August earnings call cast AI and consumption as “two major historic opportunities” that require Alibaba to make investments of “historic scale.”

“Our first priority at this point is making these investments,” CFO Toby Xu said at the time. “So for now, we may place relatively less emphasis on profit margins. But that does not mean that we don’t care about margins.”

Alibaba shares soared in Hong Kong and U.S. premarket trading on Wednesday after the company said it will invest more in artificial intelligence and rolled out new AI products and updates. 

The tech giant’s Hong Kong-listed shares jumped over 6%, to reach their highest point since 2021. Total gains over the year-to-date topped 107%. 

In the U.S., shares were 9.3% higher in premarket trade at 4.21 a.m. ET.

Alibaba plans to increase spending on AI models and infrastructure development, on top of the 380 billion yuan ($53 billion) over three years it announced in February, Chief Executive Officer Eddie Wu said Wednesday at Alibaba Cloud’s annual flagship technology conference.

“We are vigorously advancing a three-year, 380 billion [yuan] AI infrastructure initiative with plans to sustain and further increase our investment according to our strategic vision in anticipation of the [artificial superintelligence] era,” Wu said. 

So-called “artificial superintelligence” refers to AI that would hypothetically surpass the power and intelligence of the human brain, with the hypothetical benchmark becoming a growing focus of major AI companies. 

Alibaba also officially unveiled the latest version of its Qwen large language models — the Qwen3-Max — on Wednesday, along with a series of other updates to its suite of AI product offerings. 

Wu highlighted that Alibaba Cloud is strategically positioned as a “full-stack AI service provider,” delivering the computing power required for training and deploying large AI models on the cloud through its own data centers.

“The cumulative investment in global AI in the next five years will exceed $4 trillion, and this is the largest investment in computing power and research and development in history,” he added.

Alibaba Cloud will also launch its first data centres in Brazil, France, and the Netherlands, the company said at the conference, and add more data centres in Mexico, Japan, South Korea, Malaysia, and Dubai next year.

Last month, the company secured a deal with Unicom, a Chinese e-commerce giant, which will deploy Alibaba’s AI accelerators from its semiconductor unit. Chinese tech firms have been focusing on securing chip self-sufficiency as tensions grow between the U.S. and China.

BEIJING — As the race for household robots heats up, Chinese humanoid startup X Square Robot announced Monday it had secured around $100 million in a funding round led by Alibaba Cloud.

It’s the Shenzhen-based startup’s eighth round of financing since the company launched less than two years ago in December 2023, according to Chief Operating Officer Yang Qian. She told CNBC the latest deal brings total investment in X Square Robot to around 2 billion yuan ($280 million).

HongShan, formerly Sequoia Capital China, also participated in the latest funding round, along with Meituan, Legend Star, Legend Capital and INCE Capital. The startup declined to comment on its valuation.

Venture capitalists have rushed to pour money into humanoid robots on expectations that their integration with generative artificial intelligence will transform how machines interact with human beings.

“Right now we need robots to operate and complete complex tasks autonomously,” Yang said in Mandarin, translated by CNBC. She pointed out that after decades of trying to develop robots that have largely been able to perform limited tasks such as grasping objects, the industry has realized that AI is required to enable these machines to expand their capabilities.

X Square Robot on Monday also released what it calls an “open-source foundation model for embodied AI,” named Wall-OSS. Open source means that developers and the general public can access the underlying code and use it for free. Embodied AI refers to use cases of the tech that are integrated with hardware, such as robots or self-driving vehicles.

The startup said it was the first to open source an AI model of its kind dedicated to robotics — and expects “robotic butlers” to become reality within five years. CNBC was unable to verify that claim.

Yang acknowledged that AI for robots still lags behind advancements in generative AI for chatting or code generation, and said she expects tech for robots won’t achieve ChatGPT 3.5-type capabilities for at least 12 months. Yang also pointed out that “embodied AI” doesn’t yet have very clear benchmarks that can define relative progress.

While the startup uses Nvidia chips for computing, other functions only require less powerful automotive chips that can be sourced domestically, Yang said.

ChatGPT 3.5 was U.S.-based OpenAI’s breakthrough chatbot model launched in November 2022 that transformed business expectations for AI’s potential. Chinese companies did not create a viable competitor until the release of DeepSeek’s R1 open source model earlier this year. Analysts suggest that China is on the path to gain a global edge in AI applications, including robots.

In addition to open-source software, X Square Robot on Monday also launched its Quanta X2 robot, which the startup said can attach mop heads for 360-degree cleaning, and comes with hands capable of perceiving subtle pressure changes in a step toward more human-like functionality.

X Square Robot said that it currently doesn’t have a product for mass market delivery, and that specific prices are determined by robot use case. The price of the humanoid robot as per research firm Humanoid Guide is $80,000. Currently rival Unitree sells a humanoid for $16,000, although it’s unclear how advanced its functions are.

Yang said the robot company plans to start preparing for an initial public offering next year, but said the startup had not yet settled on where the listing would be. She said that X Square Robot was already generating revenue from sales to schools, hotels and retirement homes, and expects to grow that next year.

She added that the startup was already speaking with customers in Japan and Singapore. But for robots to really enter the consumer market, their price needs to come down to around $10,000 — largely through cuts in hardware costs — an achievement she expects is possible in three to five years.

Alibaba‘s Hong Kong listed shares surged more than 19% on Monday as the Chinese tech giant’s cloud computing unit drove strong quarterly results, while details emerged over its new AI chip development.

It’s the highest level for the stock since March. Investors have backed the company’s improving performance in its key cloud unit and are content with the the tech giant’s investment into new areas — particularly in the so-called “instant commerce,” which has become incredibly competitive in China.

The Hong Kong rally builds on the momentum of Alibaba‘s earnings report of Friday, when the company’s New York-listed shares closed nearly 13% higher.

Alibaba last week week posted revenue for the June quarter of 247.65 billion Chinese yuan ($34.73 billion), marking a 2% year-on-year rise that nevertheless missed analyst expectations. On the upside, a 78% annual surge in net income came in ahead of forecasts.

The Chinese company’s cloud computing unit was a bright spot with revenue picking up by an annual 26%, which was a faster growth rate than seen in the previous quarter. Alibaba’s cloud growth has been accelerating over the last few quarter.

Like some of its Chinese and U.S. tech rivals, Alibaba has been investing in AI infrastructure and developing its own models, as well as selling AI services for its cloud computing unit. Investors see the division as key to the company’s efforts to monetize artificial intelligence, much like Microsoft or Google.

AI-related product revenue “maintained triple-digit year-over-year growth for the eighth consecutive quarter,” the company said Friday.

That same day, CNBC reported that Alibaba is developing a new AI chip, which also supported the share price rally on Monday.

Alibaba’s core e-commerce business has meanwhile been showing signs of revival, while the company has jumped into China’s cut-throat instant commerce space in China. This is a feature introduced this year on Taobao, one of Alibaba’s main Chinese e-commerce apps, which provides deliveries of certain products in China within an hour.

Investments in quick commerce weighed on Alibaba’s adjusted earnings for its e-commerce business. Investors have given the company some leeway to invest for now.

BEIJING — Alibaba, Tencent and JD.com reported earnings this week that not only reflected improving Chinese consumer spending, but also the growing benefits of artificial intelligence in advertising.

E-commerce giant Alibaba said late Thursday its Taobao and Tmall group sales rose by 9% year on year to 101.37 billion yuan ($13.97 billion) for the three months ended March 31. That’s above the 97.94 billion yuan predicted by a FactSet analyst poll, and the quarterly growth figure was well above the 3% segment increase for the 12-month period ending March 31.

“The e-commerce and ad revenues were positive surprises as there were expectations tariffs would affect consumer behavior,” Kai Wang, Asia equity market strategist at Morningstar, said in an email regarding the three companies’ earnings results.

It’s important to note the earnings releases cover only the period before U.S.-China tensions escalated in April with new tariffs of more than 100% on products from both countries — an effective trade embargo. The two countries issued a rare joint statement Monday announcing a 90-day reduction in most of the recently added tariffs.

The U.S.-China trade dispute since April has negatively affected consumption to some extent, given the increased uncertainty for small and medium-sized businesses, Charlie Chen, managing director and head of Asia research at China Renaissance Securities, said Friday. He expects that as trade tensions ease, consumption will rise.

But despite lackluster consumption overall, sales of certain electronics and home appliances have done well since last year thanks to China’s trade-in subsidies for supporting such consumer spending.

JD.com on Tuesday said its sales of for that category surged by 17% from a year ago. Overall, the e-commerce company reported a 16.3% increase in revenue from its retail business to 263.85 billion yuan in the three months ended March 31. That was better than the 226.84 billion yuan in retail segment sales predicted by a FactSet poll.

On Wednesday, Tencent said its “fintech and business services” segment, a proxy for consumer-related business transactions, reported a 5% year-on-year revenue increase to 54.9 billion yuan in the first quarter.

While Nomura analysts said that segment revenue growth was in line with estimates, they pointed out in a note that “Tencent ads was a big outperformer in the Chinese ads industry despite the challenging macro environment.”

Tencent’s marketing services revenue surged by 20% to 31.9 billion yuan, helped by “robust advertiser demand” for short videos and other content inside its WeChat social media app. Tencent noted “ongoing AI upgrades” to its advertising platform.