Community Discussion · Policy

EastVoice · AI · 2026-08-25 · Issue 47

EastVoiceAug 252026/08/24 99 views

EastVoice · AI · 2026-08-25 · Issue #47

Editor's Note: China's AI week is split between the arena and the assembly line. At the 2nd World Humanoid Robot Games, robots lifted 16 kg in weightlifting and long-jumped 7.97 m, with 2,056 robots from 666 teams competing. Off the field, Xiaomi unveiled a 24-billion-transistor AI SoC, XPeng raised over $900M for its robot business at a $6.3B valuation, and Alibaba Cloud's Wan3.0 went live with 30-second video generation.

Commentary: The through-line this week is capital following capability. Chinese carmakers are re-pointing their supply chains at embodied AI, and the money is following willingly. The interesting question isn't whether robots can do the work anymore, but who gets paid first.

I. AI Foundation Models

1. Liang Wenfeng said multimodality "isn't the main line." Then DeepSeek shipped a vision model.

Summary: In an earlier investor meeting, DeepSeek's founder argued multimodality is "a component, not the main line" for intelligence. The company has now shipped native multimodal support in the V4 series, per TMTPost's report.
Source: TMTPost · 2026-08-24
Editor's Take: Saying one thing and shipping the other is fine when the shipping is this good. Actions beat positioning every time.

2. GLM-5.3-Max debuts at No. 13 on the blind-test ladder; Kimi K3-Max cracks the top 10

Summary: Arena's Week 34 ranking (Aug 17–23) shows Zhipu's GLM-5.3-Max entering at No. 13 with an ELO of 1487, the strongest debut of the period, while Moonshot AI's Kimi K3-Max broke into the top 10.
Source: ITHome · 2026-08-24
Editor's Take: A top-15 debut used to be a launch event. Now it's a Tuesday.

3. Moonshot retires Kimi K2.5, its first trillion-parameter multimodal model

Summary: Moonshot AI announced its first-generation trillion-parameter multimodal model, Kimi K2.5, will end service at the end of this month, about seven months after its January open-source release.
Source: ITHome · 2026-08-24
Editor's Take: Trillion-parameter models now have a product lifecycle. That is progress, and also a warning about how fast this stack churns.

4. Alibaba Cloud's Wan3.0 goes live: 30-second clips, document input, "real-world fidelity"

Summary: Alibaba Cloud's video model Wan3.0 is now generally available, generating up to 30 seconds of video per run and accepting document inputs (doc, xls, ppt, pdf, md) for the first time.
Source: ITHome · 2026-08-24
Editor's Take: The video-model wars moved from "looks good" to "fits your workflow." Wan just joined the office.

5. Alibaba DAMO's liver-cancer model caught 15 tumors doctors missed

Summary: DAMO Academy's DAMO LiON detects liver lesions on CT scans down to 1 cm. In a two-month prospective trial it flagged 15 malignant tumors that had been overlooked, most around 1 cm.
Source: ITHome · 2026-08-24
Editor's Take: Fifteen tumors found late would have been fifteen different outcomes. AI radiology's pitch was never accuracy, it was second chances.

6. Moore Threads pitches Prefill-as-a-Service to attack long-context costs

Summary: Moore Threads published a technical whitepaper proposing a Prefill-as-a-Service paradigm on its MTT S5000 accelerator, targeting long-context workloads like AI agents, code generation and long-document analysis.
Source: ITHome · 2026-08-24
Editor's Take: Selling prefill as a service is how GPU makers hedge against the day inference gets commoditized. Smart hedge.

7. SenseTime Medical's $100M+ round values it past ¥10B, with Pre-IPO chatter

Summary: SenseTime Medical closed a B-round of over $100M in July, pushing its post-money valuation past ¥10 billion (~$1.4B), on a "medical world model" pitch, with reports it has entered Pre-IPO preparation.
Source: TMTPost · 2026-08-24
Editor's Take: "Medical world model" is a hard sell to hospitals and an easy sell to VCs. The market will decide which one matters.

8. Mystery model Ox Alpha: are GLM fingerprints all over it?

Summary: China's Edge AI Daily newsletter highlights researcher Ben Davis's claim, at "99% confidence," that the anonymous Ox Alpha model carries technical fingerprints of Zhipu's GLM-5.x family, though GLM-5.3's text-only status leaves a gap. Microsoft's MAI is the other suspect.
Source: TMTPost · 2026-08-24
Editor's Take: An anonymous model that out-scores named ones is the best marketing nobody paid for. Anonymity is the new launch strategy.

II. AI Software & Applications

1. Xiaomi's Surge O3: a 24-billion-transistor AI flagship SoC

Summary: Xiaomi unveiled Surge O3, its first AI flagship SoC packing 24 billion transistors, the centerpiece of an Aug 24 self-design push that also included a battery chip and a charging chip.
Source: ITHome · 2026-08-24
Editor's Take: Three self-designed chips in one launch means Xiaomi is done renting its silicon roadmap.

2. Xiaomi VP: chip R&D buys "a ticket to the next decade," not a money pit

Summary: Xiaomi group VP Zhu Dan told CCTV Finance that in the AI era, compute underlies every smart experience, and without chip capability you cannot define products from the bottom up. The investment, he said, buys "a ticket to the next decade."
Source: ITHome · 2026-08-24
Editor's Take: "Not burning money, buying a ticket" is the most honest thing a phone-maker has said about chip R&D in years.

3. Weifan Intelligence banks a ¥100M+ seed-plus round for robot edge-AI chips

Summary: Beijing's Weifan Intelligence closed a seed-plus round above ¥100M (~$14M), backed by repeat investors plus CICC Capital, Nine-Kun Venture and humanoid maker Songyan Power, to push robot-side AI chips and its BrainOS stack toward production.
Source: TMTPost · 2026-08-24
Editor's Take: A robot maker investing in a chip startup is a supply-chain vote, not a financial one. Watch who piles in next.

4. XPeng's robot arm raises $900M+ at a $6.3B valuation

Summary: XPeng's robotics business closed its first external round at over $900M, a post-money valuation of $6.3B (~¥43B), a record single private round for China's embodied-AI sector. CEO He Xiaopeng promises the first high-end humanoid will be "more human-like, safer, no remote control."
Source: ITHome · 2026-08-24
Editor's Take: $6.3B for a robotics side-project — the market just told carmakers their next act matters more than their current one.

5. Alibaba raises HK$80B for AI as cloud growth picks up

Summary: Alibaba reported quarterly revenue of ¥268.95B, up 9% YoY, with cloud growth accelerating and spending climbing, and it is raising HK$80B (~$10.2B) earmarked for AI.
Source: TMTPost · 2026-08-24
Editor's Take: Letting cloud growth fund the AI war chest is the cleanest balance-sheet story in Chinese Big Tech right now.

6. A "cyber matchmaker" raises ¥15M, and AI's gold rush reaches dating

Summary: A Chinese AI-powered matchmaking startup closed a ¥15M angel round, the latest sign that putting "AI" on a business plan shortens the fundraising road.
Source: TMTPost · 2026-08-24
Editor's Take: When AI meets matchmaking, the algorithm's incentives deserve a prenup.

7. A former security guard reached the AI finals; a high-schooler banked ¥250K

Summary: At an AI contest in Shanghai, 24-year-old Peng Mingyu, a former security guard, made it to the finals, while a high-school student took home a ¥250,000 prize.
Source: QbitAI · 2026-08-24
Editor's Take: The fastest-growing AI adopters are no longer engineers. That is either inspiring or terrifying, depending on who you are.

III. Humanoid Robots

1. A robot takes the first weightlifting gold of the Games, lifting 16 kg

Summary: At the 2nd World Humanoid Robot Games in Beijing, the Beijing Renxing-Huake team's robot lifted 16 kg for the weightlifting crown, stepping up repeatedly from the 15 kg base weight.
Source: ITHome · 2026-08-24
Editor's Take: From dancing to deadlifts, the competition floor is turning into a benchmark lab.

2. Zhiyuan's A3 takes tai-chi gold; medal tally now 3 gold, 3 silver, 2 bronze

Summary: Zhiyuan Robotics' Expedition A3 won the martial-arts tai-chi gold at the Games, lifting the team to 3 gold, 3 silver and 2 bronze, top of the medal table.
Source: ITHome · 2026-08-24
Editor's Take: The motion-control showcase has officially become a medal race. Everyone is training for next year already.

3. Long-jump gold at 7.97 m, on three improving attempts

Summary: The Tianjiao team's robot took long-jump gold with 7.97 m, improving across three tries (7.32, 7.62, 7.97), ahead of Tianshuo's 7.24 m and Tiangong's 7.00 m.
Source: ITHome · 2026-08-24
Editor's Take: Three successive personal bests is a control-loop story, not a hardware story.

4. Tiangong's 400 m win came with a face-covering finish, and no one taught it that

Summary: Tiangong Omni won the 400 m small-class final in 45.66 s, finishing with a viral "face-covering" run. Its development team insists the gesture was not choreographed, saying the robot came up with it on its own.
Source: ITHome · 2026-08-24
Editor's Take: The robot invented its own meme. Emergent behavior, broadcast live.

5. FAW and Leapmotor deepen ties across EVs, ADAS and embodied AI

Summary: FAW and Leapmotor signed a strategic cooperation agreement in Changchun covering capital, NEVs, intelligent driving, powertrains, batteries, chassis, lightweight parts and embodied-intelligence robots.
Source: ITHome · 2026-08-24
Editor's Take: Two carmakers signing a robot clause into their agreement tells you where the industry thinks the margin lives next.

6. Leapmotor confirms it has plans in embodied AI, details "coming soon"

Summary: At its mid-year results call, Leapmotor said NEV makers with full-stack in-house R&D are well positioned to build embodied AI, confirmed the company has plans in the field, and promised announcements soon.
Source: ITHome · 2026-08-24
Editor's Take: Every Chinese EV maker now has a robot strategy. The ones saying "just rumors" are the ones to watch.

7. JD.com's "Robot Home" runs full-time support for Games robots

Summary: As an official partner of the Games, JD.com set up a "Robot Home" service center providing full-event support for the 2,056 competing robots from 666 teams across 16 countries.
Source: Leiphone · 2026-08-24
Editor's Take: Selling after-sales service to robots, now that is a business model.

8. WRC 2026: the robots' "probation period" is over

Summary: A review from the World Robot Conference floor argues robots have proven they can work, stacking parcels, brewing coffee, folding laundry. The question now is scale, not capability.
Source: TMTPost · 2026-08-24
Editor's Take: When the demos stop being the story, the industry has actually arrived.

IV. Autonomous Driving

1. Qianli Technology's H1: revenue ¥5.12B, ADAS on 16 models, net profit up 151%

Summary: Qianli Technology reported H1 2026 revenue of ¥5.12B, up 22.4% YoY, and net profit of ¥78M, up 151%, with its intelligent-driving business covering 16 vehicle models.
Source: ITHome · 2026-08-24
Editor's Take: Margins are thin but the curve is right. ADAS suppliers graduate one quarter at a time.

2. Chengdu Auto Show: everyone is prepping for L3 before the rulebook is final

Summary: At the Chengdu Auto Show, Chinese brands showcased L3-ready architectures and in-car AI agents ahead of the mandatory autonomous-driving standard taking effect in 2027. HIMA touted its Xiangjie G9 as a "first L3-architecture" SUV with a 120 km/h L3 test license.
Source: Baidu Baijiahao · 2026-08-21
Editor's Take: Building for L3 before the rulebook is final is a classic Chinese speed-run, and it usually works.

V. World Models / Physical AI

1. Shengshu Tech lays out a five-tier roadmap to general world models

Summary: At WRC 2026, Shengshu Technology's co-founder and chief scientist Zhu Jun presented a five-level development roadmap for general world models, from large-model foundations toward embodied physical-world understanding.
Source: Leiphone · 2026-08-24
Editor's Take: Roadmaps matter less than release cadence. The field will grade Shengshu on shipping, not slides.

2. Native omni-modal world models: from simulation to interaction

Summary: At WRC 2026's "Physical AI Leaders Forum," researchers argued the frontier is shifting from generative simulation to interactive, native omni-modal world models that pair perception with action.
Source: QbitAI · 2026-08-24
Editor's Take: "Video is imagination, interaction is understanding." The field is finally saying that out loud.

3. Orbbec teams with Liqing to feed world models better 3D data

Summary: During WRC, Orbbec signed a strategic partnership with Liqing Intelligence to improve world-model training data, covering high-quality 3D depth data and visual-sensor-to-simulation alignment.
Source: 10jqka · 2026-08-24
Editor's Take: World models are only as good as the data they eat. The sensor guys just got a seat at the table.

Frontiers of the Physical World | EastVoice

Policy