Today’s CTR: China tech spent the past 24 hours doing what it increasingly does best: turning frontier ideas into financing events, platform wars and industrial policy in miniature. Artificial intelligence [AI] video is suddenly valuable enough to attract sovereign-scale cheques. Office software is being rebuilt around agents, not calendars. Robotics is marching from viral demos toward public markets. Even gaming news came with a reminder that China’s consumer internet still knows how to scale entertainment across devices. The mood is not exuberant, exactly. It is more like disciplined impatience: capital wants proof, platforms want control, and everyone wants a bigger compute bill yesterday.
Kuaishou’s Kling AI raises nearly $3 billion, because video models have become the new datacenter rent collector
Kuaishou secured 19 billion yuan, or about $2.8 billion, for Kling AI, its artificial intelligence [AI] video-generation unit, in a round that values the business at $15 billion before the investment. The company is still seeking to raise roughly $3 billion in total for a 16.7% stake, with Alibaba, Tencent and Baidu among the investors, while Kuaishou is expected to retain control.
The deal is notable less because China needed another AI financing headline than because this one has a clearer commercial thesis. Generative video burns cash quickly, but it also sits closer to advertising, entertainment and creator tools than many general-purpose model businesses. That makes Kling a useful test of whether Chinese AI can escape the “impressive demo, awkward invoice” trap.
The uncomfortable detail is cost: Kling reportedly posted net losses in 2024 and 2025 as computing expenses climbed. Investors are not funding a software company so much as a race to convert expensive inference into recurring demand.
Closing thought: Kling’s round says China’s AI-video market has entered its blockbuster era, but the sequel will be about margins. Source
Alibaba folds its AI work apps into one platform as the enterprise-agent fight gets serious
Alibaba is consolidating its AI productivity applications into a single enterprise platform after a leadership shake-up at DingTalk, its workplace communication unit. The move is meant to sharpen its response to Tencent and ByteDance in China’s increasingly competitive AI office software market.
This is a defensive reorganization with offensive intent. Alibaba cannot afford to let DingTalk become yesterday’s enterprise chat app while rivals define the next layer of workplace software through AI agents that write, search, summarize and transact. In this market, the operating system is not the device; it is the workflow.
Caixin’s digest notes that Tencent’s Workbuddy had 8.85 million monthly visits in March, while DingTalk’s Wukong had 530,000. That gap explains the urgency better than any strategy memo could.
Closing thought: Alibaba is trying to turn a product sprawl problem into a platform story before Tencent turns the office into another super-app battlefield. Source
Unitree wins approval for a $618 million STAR Market IPO, giving China’s robot boom a public-market scoreboard
Unitree Robotics received approval from China’s securities regulator for an initial public offering [IPO] on Shanghai’s STAR Market. The company plans to raise 4.2 billion yuan, or about $618 million, for research, development and production, with a potential valuation of about 40 billion yuan.
The approval gives investors a cleaner way to price China’s embodied-intelligence story. Unitree has already become the most visible Chinese robotics name abroad, thanks to quadruped and humanoid robots that move well enough for viral clips and cheaply enough to attract labs, universities and developers. The IPO will test whether that attention can translate into durable orders.
The broader reach is industrial. China wants robots not merely as gadgets, but as labor, factory infrastructure and strategic hardware. Public funding helps push the sector from demonstration floors into supply chains, which is where the real money, and the real engineering pain, begins.
Closing thought: Unitree’s listing is less a finish line than a quarterly-reporting machine for China’s humanoid ambitions. Source
Tencent puts AI body tracking into Just Dance: Party, a small game launch with a larger interface lesson
Tencent launched Just Dance: Party, a Ubisoft-licensed mobile rhythm game, across iOS, Android and HarmonyOS devices. The game uses Tencent’s proprietary AI-powered skeletal-tracking algorithm to capture full-body motion through a smartphone’s front-facing camera, removing the need for dance mats, motion controllers or other accessories.
On its face, this is consumer-entertainment news. Underneath, it is another example of AI moving quietly into interaction design, where the winning feature is not a chatbot but the disappearance of hardware friction. A phone camera becomes the controller; the user becomes the peripheral.
The product also shows Tencent’s enduring advantage in licensed content, game distribution and social play. In a market where regulators and consumers have both made pure gaming growth harder, AI-enabled formats offer a way to refresh old franchises without pretending every title must become the metaverse.
Closing thought: Tencent’s dance game will not redraw China tech, but it neatly shows how AI becomes mainstream: first as a feature nobody has to think about. Source
Party Animals heads to mobile, and China’s games industry remembers the obvious: distribution still matters
Recreate Games announced at the 2026 TapTap Game Conference that it is developing a mobile version of Party Animals, its physics-based multiplayer hit. The studio did not provide a testing schedule or release window, and jokingly said the “mobile version folder has been created,” which is either admirable candor or the most honest product roadmap in gaming.
The move is strategically straightforward. Party Animals built a strong reputation on PC and Xbox, earning a Very Positive rating on Steam and nominations at The Game Awards and the BAFTA Games Awards. Bringing that brand to mobile opens a much larger Chinese and global addressable market, provided the team can preserve the game’s chaotic physics and social feel on touchscreens.
The risk is that mobile is not merely another screen; it is a different business model, control scheme and attention economy. A lazy port would squander a charming franchise. A polished one could turn a cult multiplayer title into a much broader live-service property.
Closing thought: Party Animals on mobile is early, but the logic is mature: China’s best game studios increasingly see global intellectual property [IP] as something that must travel across platforms. Source
