RMB 289 a day on one page, RMB 2,000 on the same app: five products, four cost lines, and one price that is not settled yet.

TwinDEX ran a 24-step chemistry experiment with no real-robot teleoperation data in post-training. Here is what 2026 repriced in robot training data.

China's robot IPO window narrowed in H2 2026: 14 companies listed, a STAR Market applicant withdrew, and the test became who pays the revenue.

Chinese robot hands now cost a fraction of their 2025 prices, yet fewer than 1% of them work on automotive lines. The industry has three separate ways of judging a hand, and the one that measures lifespan has no common scale.

XPeng's humanoid robot walked off its own automated line on September 8, 2026. Who is building China's robots, and can the EV supply chain deliver them?

Global humanoid production passed 20,000 units in 2025, ten times 2024. Factory deployments stay case-by-case. A four-layer data check, from Longcheer's 64,828 tasks to Unitree's IPO filing.

Chinese makers built over 97% of the world's humanoid robots in H1 2026 but exported only 12.4%. What IFA 2026 shows about Europe's barriers.

Unitree lost RMB 230 billion in 13 days, yet still trades 251.7% above issue. Why the anchor held, and six flags to check.

Led by China with eight countries, ISO 18646-5 is the first global benchmark for legged robots. Why legs, why now, and what this ruler cannot measure.

A $399 duck: $1M in 7 hours, 10,000 pre-orders in 5 days, a Chinese chip stock at its limit-up. How Microduck redrew the physical AI entry ticket.

AGIBOT shipped the most humanoid robots and won 46 medals, yet carries roughly one-sixth of Unitree's value. Inside the mismatch.

Who took gold at Beijing's second World Humanoid Robot Games: a roundup of the winning robots and what the scores reveal.

YOUIBOT ranks first in industrial mobile manipulation robots at 6.1% share. Its three market denominators span RMB 4.2bn to RMB 82bn.

HokMind, a Beijing startup from March 2026, sells a bionic face with 32 degrees of freedom. This report sizes the market and separates three layers of evidence.

Agile Robots booked EUR 300m of revenue in 2025 and guided to EUR 600m for 2026, yet 85 to 90 percent of that growth comes from business the company already had, because the firms it bought supply only 10 to 15 percent of this year's new orders. Those deals bought doors rather than turnover: an insolvent Franka Emika, BMW's idealworks, thyssenkrupp's automation unit, BAER in Gemmingen, audEERING in Munich, and XNG in Bengaluru. This file opens each door, asks how far a force-control company can ride an integration business, and lists the three disclosures it has never made. Article structure: 1. The verdict 2. Why this market exists 3. The company 4. Products 5. Customers 6. Supply chain 7. Competition 8. Money 9. The moat 10. Risk register 11. What to watch

On September 20, a Chinese company put two personal robots on sale in China, starting at RMB 19,999 (about $2,800). Anyone typing "a robot that does my chores" into a search box now has, for the first time, an option they can actually order. It does not do chores. The one that does sits on the other side of the Pacific, costs roughly seven times as much, and still keeps a remote operator on call for the hard tasks. Article structure: - 1. What People Search For, and What Is Actually on Sale - 2. What $2,800 Buys - 3. The Price Tag on Chores - 4. Why Companionship Comes First - 5. What to Watch Over the Next Twelve Months

On Sept 16, 2026, a Tesla team arrived in Ningbo, a port city in Zhejiang province, and began a mass-production audit of its robot suppliers the next day. The names reported from that round, Tuopu Group, Sanhua, Joyson Electronics, Zhejiang Rongtai and Xinquan, are almost all long-standing suppliers to Tesla's car business. Why would a robot have to clear the production lines of auto parts makers first? Because what an audit measures is cycle time, yield and cost, and those three sit with companies the car industry has spent decades hardening. Tesla's own description of the task is that "almost every part is new, with no ready supply chain." The list is therefore more than a purchase order. It is Tesla looking for a manufacturing path it can repeat. Article structure: - 1. The audit: four departments on site, three answers - 2. The list: why it is almost all auto parts makers - 3. The threshold: how far an audit is from mass production

On July 8, 2026, engineers and surgeons at the University of California San Diego published a paper in Nature reporting that two teleoperated humanoid robots had performed laparoscopic gallbladder removals on live pigs. The platform was an off-the-shelf Unitree G1, nicknamed Surgie, standing five feet tall and weighing 60 pounds. The dedicated surgical system it was measured against weighs about 1,800 pounds. The paper also supplies the numbers that matter: straight-line tracking error of about 1.3 millimeters, circular tracking error of about 10.4 millimeters, and 156 milliseconds of latency. Across two dry-lab tasks the humanoid already beat unassisted human hands, while remaining clearly behind the da Vinci Xi. This article unpacks that scorecard, and the four gates still standing between the robot and a clinical operating room. **Article structure** : - 1. The Surgical Record: Two Live Porcine Cholecystectomies - 2. Why a General-Purpose Humanoid: Footprint, Reach and Cost - 3. Measuring Precision: Straight Lines, Curves and 156 Milliseconds - 4. Four Gates Before Clinical Use: Sterility, Ports, Recalibration and Remote Operation - 5. Conclusion: What the First Time Proves, and What It Does Not

On Sept. 17, D-Robotics said it raised $400 million in a Series C round led by Mirae Asset of South Korea, its third round this year. The company does not build robots. It sells the chips that go inside them. The announcement carried three numbers: $400 million raised, 8 million chips shipped to date, and coverage of more than 50% of customers in embodied AI. The three numbers describe the same business, yet they point in different directions. The announcement does not say which machines took those 8 million chips, or how the 50% was calculated. Article structure: - 1. $400 Million: A Third Round This Year, No Valuation Disclosed - 2. 8 Million Chips: Where They Went - 3. "Over 50% Coverage": One Metric, Three Versions - 4. Spun Out of Horizon Robotics: Engineering Lineage and Product Boundaries - 5. Conclusion: Revenue Structure and Three Metrics to Watch