MiniMax positions it as an end-to-end coding model that can take a software task from diagnosis through delivery.

More than 3,000 guests are expected to attend, including industry executives and technology experts from around the world.

Alibaba has reduced prices across its Qwen-Audio voice models by as much as 95%, including a roughly 70% cut for text-to-speech services.

China Life has increasingly used both S transactions and fund investments to gain exposure to technology assets.

The roughly 3-meter-tall robot weighs about 500 kilograms with a rider and can switch between bipedal and quadrupedal modes.

For Momenta, the partnership also provides a route for its driving technology to reach overseas markets through Stellantis’ global brands.

Nio and Geely began working together on battery swapping in November 2023, when Geely joined Nio’s battery-swapping alliance.

CXMT founder and memory titan GigaDevice’s venture arm reportedly behind the AI champion with a 5 billion yuan bet.

He has been involved in the development of Qwen’s core language models as well as Qwen-Coder and multimodal models.

Many teachers and students are already using AI agents for lesson preparation, research and learning, but mostly on an individual basis.

The 104 Zhejiang companies making the list span internet services, automotive, chemicals, advanced materials and smart manufacturing.

Its Qwen team plans to train a new model with 5 trillion to 10 trillion parameters as it moves toward artificial superintelligence, or ASI.

The platform’s customer base had grown sixfold over the past year as AI agents began moving into core business processes.

Users can activate QwenWork with a single voice command and have the AI turn conversations into tasks, notes and follow-ups.

The V900 is scheduled to enter mass production in the first quarter of 2027 and will be deployed across Alibaba Cloud data centers.

This move comes as the Hangzhou-based tech giant ramps up AI infrastructure investment to meet surging demand for computing power.

The restructuring comes four days after Ant adopted Alibaba’s QwenWork as its AI workplace platform for all employees.

The new model increases the number of degrees of freedom while reducing the overall size and improving drive precision, according to Unitree.

CXMT said the G5 yield ramp-up will take time and that new products must go through customer testing and certification before broader adoption.

As AI computing drives server power densities toward the megawatt range, conventional air cooling is becoming increasingly impractical.

Tesla’s reported supplier strategy highlights the role of China’s existing automotive manufacturing base in its robotics push.

The rebound reversed a monthlong slide that had taken the stock below the 500-yuan mark for the first time on September 10.

For Chinese AI companies expanding overseas, the partnership marks a shift in how their products can reach foreign consumers.

A supernode essentially connects thousands of computing cards through high-speed links so they can work together as a single computing system.

The Hangzhou experiment takes a different approach, positioning seniors as AI users, creators and contributors.

The app then limits access to four learning functions: photo-based Q&A, AI writing, translation and deep research.

The latest additions broaden that effort from drug and product information to medical knowledge and nutrition data.

Through its partnership with Stellantis, Leapmotor International now has more than 1,000 sales and service outlets across 45 markets.

The essay offers an unusual glimpse into how AI automation is being perceived by engineers at the cutting edge of China’s AI industry.

The first vehicle rolling off the line less than a year after the partnership was signed highlights the speed of the approach.

Reuters, Airbnb and Pinterest have also adopted Qwen models as companies seek to reduce their reliance on proprietary AI systems.

Huawei says the 3D architecture can halve the time required for mechanical, electrical and plumbing delivery.

The appointment would mark one of DeepSeek’s most significant senior hires outside its core tech team since the firm was founded.

The strategy also marks a potential shift for China’s leading robotics companies from exporting products to exporting ecosystems.

Whether Lei’s latest visit was simply a high-profile show of support or a sign of deeper industrial cooperation remains unclear.

The original G1 was launched in 2024 at 85,000 yuan and helped Unitree become the world’s top humanoid robot maker by shipments in 2025.

Han said this could change the starting point of commerce from standardized products and fixed app pages to user intent.

For developers, the bigger question is whether DeepSeek will bring its aggressive pricing strategy to voice AI.

AI agents could make services that once relied on scarce experts or were too costly to deliver at scale available to a much wider population.

This move comes as the Chinese robotics maker seeks to push deeper into the underlying software stack for embodied AI.

A new “Health” section has been added to Afu’s homepage, alongside its AI chat and medical-service entrances.

As an official DeepSeek partner, Tencent’s WorkBuddy has fully integrated V4.1 Flash in China and is offering a two-week promotional discount.

The potential listing comes after China expanded the STAR Market’s so-called fifth listing standard in June to cover AI companies.

Rather than simply discussing AI in theory, the guide provides step-by-step instructions and ready-made prompts for each scenario.

He expects that “ChatGPT moment” in embodied intelligence to arrive in as little as two to three years, though it could take five to 10 years.

He also criticized what he called overly aggressive innovation, in which insufficiently tested designs are incorporated into vehicles.

The company expects the technology eventually to expand beyond flagship smartphones into smart vehicles, edge servers and wearables.

DeepSeek’s latest strategy combines aggressive pricing to win users and peak/off-peak rates to encourage better use of computing capacity.

More revealingly, the company’s internal testing questionnaire asks users whether the model could fully replace the current DeepSeek V4 Pro online.

The launch reflects a shift in AI-agent development from helping individuals work faster toward assigning AI agents specific roles.

That puts Deli up against an increasingly concentrated market led by Shenzhen companies including Bambu Lab, Creality and Anycubic.

The hiring points to DeepSeek’s growing focus on its in-house elastic computing platform, DSec, or DeepSeek Elastic Compute.

Unitree said the model overcomes bottlenecks in real-time planning, decision-making and dynamic interaction.

The model differs from many overseas AI finance products, which are generally built within individual platforms.

The Stock Connect inclusion gives Manycore access to a broader pool of Chinese investors and could improve trading liquidity.

iFlytek says it was trained and deployed entirely on domestic computing infrastructure and is now available through its open platform.

The pitch is straightforward: companies no longer need to buy a dedicated SaaS product or commission custom software for every workflow.

The rapid growth of QwenWork comes as China’s technology giants increasingly focus their AI firepower on workplace agents.

Tmall is turning those services into standardized retail products that can be bought alongside more familiar digital goods.

The model is available through Alibaba’s Qwen AI platform as an API and has been integrated into Qwen Office, Qoder and the Qwen app.

The launch comes as the AI industry moves toward putting increasingly capable models directly onto consumer electronics and machines.

It introduces four expert-level AI agents spanning key verification stages including software and hardware simulation.

The attention reflects the pressures facing an industry that has attracted large amounts of capital and talent since the rise of generative AI in 2023.

Their entry marks a rare convergence of internet and automotive capital around China’s commercial satellite industry.

Founded in 2018, Enflame develops cloud AI chips and has built four generations of architectures and five AI accelerators.

For an AI short drama, users provide the story concept while the agents collaborate sequentially to turn it into a finished video.

The security-camera maker posted 7.90 billion yuan in net profit, up 39.57%, marking one of its strongest growth rates in nearly a decade.

The move follows a high-profile visit by Serbian President Aleksandar Vučić to Minth’s future factory in Jiaxing on May 27.

The company aims to move 3D content creation from labor-intensive manual modeling toward automated, large-scale production.

That makes the program resemble an early-stage technology grant or angel investment more than a conventional corporate recruitment scheme.

The international launch comes as AI agents move beyond chatbots toward software that can independently execute multi-step tasks.

The model, unveiled Tuesday night, uses a new “Next” architecture with 125 billion total parameters but activates just 6 billion at a time.

For investors, the move offers a different perspective at a time when founder share sales are closely watched.

Insider buying can demonstrate conviction; it cannot by itself prove that Alibaba’s AI investments will generate attractive returns.

The market’s reaction reflects the tension between Alibaba’s long-term AI strategy and its near-term financial burden.

This helps to address a persistent problem with AI-generated video: maintaining continuity across a sequence.

Under the revised rules, effective from midnight Beijing time on Aug. 23, DeepSeek no longer distinguishes between peak and off-peak hours on weekends.

The fundraising comes as Alibaba enters an AI investment window defined by accelerating revenue but mounting pressure on profits.

The real test for the succession will be whether “One Geely” can deliver more than the founder alone could.

The company is using price incentives to steer non-urgent workloads toward nights, weekends and holidays

This entity expands its presence in China’s mainland as AI-driven demand for data-center power and cooling infrastructure accelerates.

Pro output tokens are therefore three times more expensive, while cache-miss input is also priced at three times the Flash rate.

The M890 supernode is also the first supernode architecture in China reported to have successfully run AI models exceeding 2 trillion parameters.

Zhipu’s expanding Hangzhou presence reflects a broader trend among China’s leading AI companies: building research, talent and commercial ecosystems outside Beijing.

The move comes just one week after Alibaba opened public testing of QwenWork, its enterprise-focused AI productivity platform.

For AI companies, owning or controlling data centers, power resources and physical capacity can translate into lower costs and greater supply certainty.

MSCI announced on July 28 that CXMT had met the eligibility criteria, with the inclusion taking effect on August 10.

The company has ranked No. 1 in China’s new-energy commercial vehicle market for four consecutive years since entering the sector a decade ago.

The architecture, known as CUBE 5.0, raises the modularization rate of the five major systems from about 30% in earlier designs to 90%.

The first wave of partners covers more than 10 sectors, including logistics, housing, local services, wealth management and automobiles.

Wan3.0 reflects a broader shift in AI video generation — from experimental content creation toward practical production workflows.

If prototype testing meets expectations, its domestically developed LPDDR6 chips could enter large-scale production within the year.

More than 2,000 global leaders and industry experts are expected to participate in discussions throughout the event.

Cainiao said that as 72-hour door-to-door delivery becomes accessible to more sellers, global e-commerce competition will enter a new phase.

The Shanghai headquarters project represents Tsingshan’s latest effort to strengthen its presence in one of the world’s key trade and financial centers.

Qwen3.8 arrives as the AI industry moves from conversational models toward agentic systems that can independently complete multi-step tasks.

QwenWork combines three previously separate products — QoderWork, MuleRun and Wukong — into a single AI office platform.

CFMoto went head-to-head with established global motorcycle names including Yamaha, KTM, Honda and BMW.

The release of V4-Flash once again challenges the industry belief that larger models automatically deliver superior performance.

The company has built a product lineup spanning four series — A, B, C and D — covering price ranges from 60,000 yuan ($8,888) to 300,000 yuan.