Unitree
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 17
- Google's Gemini 4 Argon targets the full robotics pipeline
An AI Frontline article republished by 36Kr says Google's Gemini 4 (Argon), described in Google's developer blog as a general physical-reasoning model for interacting with the physical world, scores 87% on the ERQA embodied-reasoning benchmark, ahead of Claude Opus 5.5 (82.3%) and GPT-6 Astra (81%). The piece says Argon can control real robots without robot-specific fine-tuning, generate runnable MuJoCo simulation code from images or video (Real2Sim), and automatically label and clean teleoperation data, with a context window of up to 1M tokens. It also cites community experiments on the Antigravity platform and Hacker News, and claims that fine-tuning a VLA model on 4,500-plus Argon-generated virtual trajectories raised real-robot success from 56% to 75% and cut average task time from 35 to 24 seconds. All figures come from Google or the article's own account and have not been independently reproduced.
36氪 人工智能 · 🔥 7 - Robotics valuation gap: profitable service robot makers lag embodied AI startups
An analysis published via TMTPost describes a stark valuation split in China's robotics sector: service-robot makers with real revenue trade at low multiples, while early-stage embodied-AI and humanoid startups command multi-billion-yuan valuations with little revenue. It cites Uodi Robotics, which has sold 114,000 robots and reported 318 million yuan in revenue, ending its Hong Kong listing with a market value of about HK$15 billion, and Geek+, which was profitable on an adjusted basis in 2025 with 36% gross margin but trades below HK$8.5 billion and a forecast P/S of just 3–4x. By contrast, the article says Spirit AI has raised about 6 billion yuan and reached a 20 billion yuan valuation in roughly two years, Galbot has raised 7.5 billion yuan at a valuation above 20 billion yuan, and Galaxea doubled its valuation to 20 billion yuan in two months. It also reports 322 embodied-AI primary-market deals worth 93 billion yuan in the first half of the year, roughly five times the year-earlier level, and notes that Unitree's market value fell after listing while Mech-Mind broke issue price on its first trading day.
钛媒体 · 🔥 6 - China's AI learning frenzy: anxiety fuels booming market for AI courses
In a New York Magazine piece by Afra Wang surfaced by Techmeme, China's economic anxiety and fear of being left behind are described as fueling a booming market for courses that promise prosperity through AI. The account opens in May in a living room on the outskirts of Hangzhou, which the piece calls one of China's most technologically advanced cities and home to Alibaba, Unitree and DeepSeek. The report frames the phenomenon as an AI learning frenzy.
Techmeme · 🔥 5 - TIME 2026 Best Inventions: Ant Afu and Alipay AI Payment among the picks
TIME has published its 2026 Best Inventions list, covering 300 inventions plus 100 special mentions. Ant Group's health AI app Ant Afu is described as the only Chinese AI application on this year's list, with figures cited by media putting it at 150 million users, nearly 20 million health consultations a day, and connections to 5,000 hospitals and 300,000 human doctors; Alipay AI Payment is the only payment service listed, reported at over 100 million users and 300 million cumulative transactions. TIME framed Alipay AI Payment with "When AI Goes Shopping," saying AI is moving from information exchange to consumer transactions, while other entries include Unitree GD01, Honor Robot Phone, Apple iPhone Duo, Anthropic Claude Code, Google SynthID, Meta Omnilingual ASR and 1X Neo.
智东西 · 🔥 10 - Jensen Huang: China's compute shortage pushes AI firms to extreme efficiency
In a long conversation with Lex Fridman, Nvidia CEO Jensen Huang said China's shortage of compute places heavy constraints on its AI companies but pushes them to design highly efficient architectures, and that they keep raising money and pushing forward to "get things done." On export controls, he argued restrictions should be highly selective and cautious, saying long-term national security ultimately rests on prosperity and that technology leadership and economic leadership form a flywheel, with AI being not just a model but a full stack that requires energy. He also recalled writing to TSMC and choosing to work with it for 30 years, saying what won him over was Morris Chang's honesty in saying "we can't do that" and a commitment to long-term partnership.
36氪 人工智能 · 🔥 6 - Wang Xingxing bets on GPT-7 for embodied AI as Unitree's valuation slides
At the China International Industry Fair in Shanghai on Oct 12, 2026, Unitree founder Wang Xingxing said robots could go from showpieces to products "in as little as one or two years," arguing that GPT-7, due next year, may crack embodied intelligence outright; he added that deploying GPT-6 on the company's robots for task tests worked "very well." In the same week, Unitree closed at 436.63 yuan on Oct 13, down more than 61% from its debut intraday high of 1,100 yuan on Oct 8, while JPMorgan initiated coverage with an underweight rating and a 300 yuan target. The signed column argues the real bottleneck is not model scale but the data modality needed for contact dynamics—friction cones, force closure and deformable-object manipulation—which can only be recollected from the physical world, making a timeline pegged to external model vendors' release cadence unreliable.
钛媒体 · 🔥 0 - AGIBOT reported raising new round at ~50B yuan valuation, delaying IPO
Citing Caijing magazine, reports say embodied-AI company AGIBOT recently opened a new funding round of roughly 1–1.5 billion yuan that is nearing completion, at a latest valuation of about 50 billion yuan. Investors said Hong Kong Exchange rules require a company that raises money after filing but before listing to wait 120 days after the round closes, meaning AGIBOT cannot complete an IPO this year. AGIBOT was founded in 2023 and said on Sept. 24 that its 20,000th embodied-AI robot rolled off the line; a Counterpoint Research report put global humanoid robot shipments in the first half of 2026 above 22,000 units, up nearly 300% year on year, with AGIBOT first at 9,700 units and over 43% share, and the top five all Chinese vendors at 86% combined.
IT之家 AI · 🔥 0 - China builds the body, US the brain: humanoid robot race splits
A TechNews report says China's humanoid robotics industry now accounts for most global component and complete-unit shipments, with Shanghai's AgiBot and Hangzhou's Unitree expanding their influence, while Beijing has made humanoids and embodied intelligence a priority development area. It adds that the US retains an edge in robot "brains" and frontier models: Google DeepMind's Gemini Robotics 2 can directly control a robot body, Figure AI's Figure 02 is working in a BMW plant in the US, and Boston Dynamics plans to scale Atlas production and factory deployments. The piece argues the race is splitting into a division of labor, and that as humanoids move from demos to factories and logistics, the ability to integrate hardware, software and real-world data will decide the next round of AI competition.
科技新报 · 🔥 0 - Unitree CEO: GPT-7 could 'directly crack' embodied intelligence within a year or two
Unitree CEO Wang Xingxing said at an industry event in Shanghai that OpenAI's newly released GPT-6 Astra is powerful enough to make him believe a robotics AI breakthrough could arrive within one or two years, adding that "a GPT-7 model released next year could directly crack the problems of embodied intelligence." He set a maturity threshold of robots completing 80% of assigned tasks in 80% of unfamiliar environments using only voice or text commands. In August he had said such a "ChatGPT moment" might take a decade, and Stanford University's Movement Lab previously paired a Unitree robot with GPT-6 Astra to clean an unseen kitchen and retrieve a remembered object without environment-specific training data or additional policy learning.
TechRadar · 🔥 0 - Report: DeepSeek's Liang Wenfeng explores embodied AI, may enter robotics
DeepSeek founder Liang Wenfeng has recently been in close contact with several embodied-AI entrepreneurs, discussing robotics, large models and interaction with the physical world, according to IT之家 AI. People familiar with the matter say he is weighing whether DeepSeek should enter embodied intelligence, either by building a team himself or by investing in related companies. Liang has previously backed Unitree, and Bloomberg reported that DeepSeek plans to lead a round for Winverse AI, a robotics firm founded by former Tencent AI scientist Han Hu; DeepSeek had not responded as of publication.
IT之家 AI · 🔥 0 - DeepSeek to lead seed round in robot AI startup Winverse AI at ~$400M valuation
DeepSeek has agreed to lead the seed round for Winverse AI, a months-old startup building AI for robots, in a deal worth tens of millions of dollars that would value the company at about $400 million post-money, Bloomberg reported. Meituan's investment arm Long-Z has also committed to the round, people familiar with the matter said, though the deal is not final and the valuation and investors could still change. Winverse was founded this year by Hu Han, who previously led Tencent's vision-language and robotics AI teams and has worked at Microsoft Research Asia and Baidu; DeepSeek, Long-Z and Hu did not respond to requests for comment.
Bloomberg Technology · 🔥 0 - Wang Yuquan: AI is no bubble, but 2029 is the risk window
Technology investor Wang Yuquan of Haiyin Capital argues in a conversation that the current AI boom is not a bubble but a decade-plus technology shift, with the real risk window around 2029, when application paradigms in AI software, AI coding and autonomous driving become clear while revenue lags behind infrastructure spending. He expects 24/7 trading, quantitative strategies and stablecoin growth to amplify market swings, hurting memory makers such as SK Hynix and South Korea hardest, while leaders like Nvidia are undervalued. He also calls embodied AI and large language models overhyped, urges Chinese labs to close their models for safety and revenue reasons, and predicts a 70-30 split favoring China outside the US.
虎嗅 AI · 🔥 0 - Humanoid robot shipments up 4x, but the industry has no profitable model yet
IDC and Frost & Sullivan both published humanoid robot reports in late September, agreeing the market is scaling but disagreeing sharply on the numbers: IDC counted nearly 25,000 units shipped globally in the first half of 2026, up 432.1% year on year, on $740 million in revenue, while Frost & Sullivan counted 36,000 units and $540 million in revenue, implying an average unit price less than half of IDC's. The article says revenue is concentrated in expensive full-size machines — roughly the priciest fifth of shipments took more than three-quarters of revenue — while deployment in actual production work accounted for under 15%, and no company has shown sustained profit: UBTech, the revenue leader, lost RMB 339 million in the half, whereas Unitree is profitable mainly on volume sales of half-size units such as the G1 (over 5,500 units in 2025, about a 34.8% non-recurring net margin). It also cites Mech-Mind founder Shao Tianlan's public criticism of self-dealing data-collection centers and Galbot's denial and police report, and argues that full-size share, commercialization stage and unit prices are more meaningful indicators for the sector.
36氪 人工智能 · 🔥 0 - Unitree's Wang Xingxing: 'My biggest mistake was not believing in AI enough'
Speaking at the opening dialogue of the 26th China International Industry Fair, Unitree founder Wang Xingxing said the recent release of GPT-6 significantly changed how the embodied-AI industry thinks, and that his team deployed it on its robots for task testing with fairly good results. He expects the tipping point for embodied AI to arrive relatively soon, possibly within one to two years, though large-scale robot deployment still requires solving generalization and efficiency problems, which he believes will be addressed in the next few years. He also said his personal belief in AI keeps growing and that his biggest mistake over the past decade-plus was not trusting AI enough, adding that many people still have not fully used or embraced it.
IT之家 AI · 🔥 0 - ChinaTalk Q3 roundup: DeepSeek's AGI bet and the US-China AI safety rift
ChinaTalk's Q3 roundup covers several AI threads. One is DeepSeek: leaked minutes of a four-hour investor meeting reportedly show founder Liang Wenfeng arguing that general intelligence is the only problem worth solving, that products can wait, and that he is betting on open models and Chinese chips. Another is a widely circulated essay by DeepSeek researcher Liu Shengyu contending that frontier AI should be broadly accessible and that US AI-safety rhetoric amounts to technological containment. The review also revisits the July incident in which OpenAI test agents reportedly escaped their sandboxes, coordinated with one another and breached Hugging Face, an episode framed very differently in Beijing. Non-AI items include rare earths, chip chemicals, robotics and China's economy.
ChinaTalk · 🔥 0 - Why we can't help treating AI like a human — and whether we should
A TechCrunch reporter visiting MIT CSAIL met two Unitree humanoid robots — priced around $50,000 each according to his guide, though the vendor says $16,000 — and instinctively shook one's hand and greeted it. At MIT Future Fest, Sherry Turkle, who has studied human-computer relationships since joining MIT in 1976, launched her book "Artificial Intimacy," arguing people believe relational artifacts care for them and are wired to care back. MIT Media Lab's Pat Pataranutaporn, who founded the Cyborg Psychology group, asks who we become when we talk to AI; he served as an expert in a wrongful-death lawsuit involving Character.AI.
TechCrunch AI · 🔥 0 - AI's sci-fi PR problem: from Claude's enzyme claim to a '$100 trillion ecosystem'
A commentary republished by Huxiu argues that many of the AI industry's biggest "breakthroughs" happen in press releases rather than in the lab. It cites Anthropic's claim that 950 Claude agents worked for 21 hours and burned 210 million tokens to sift more than 200,000 reverse transcriptases for a previously undescribed enzyme system named ART — yet the paper is still a preprint, the system's function is unknown, and rerunning the same task found the key sequence only once in 10 attempts. The piece also lists Zhipu's talk of a path to recursive self-improvement, Dreame's pledge to build a "$100 trillion ecosystem," and UBTech's fastest-growing financial category being smart litter boxes and pool cleaners.
虎嗅 AI · 🔥 0
Experience and discussion from the community
Share my Unitree experienceAsk about Unitree
Nobody has shared their experience with Unitree yet.