Skip to main content
Back to News
Tashi Zhihang (它石智航) Raises $455M Pre-A to Build Brain-First Embodied AI
Funding
2 min read
CN

Tashi Zhihang (它石智航) Raises $455M Pre-A to Build Brain-First Embodied AI

The AMW Read

The $455M Pre-A round is a record-breaking mega-round for Chinese embodied AI, signaling a structural shift in capital toward 'brain-first' full-stack foundation models in robotics.
NoveltySignificance
Robotics · Player Map

Tashi Zhihang (它石智航) Raises $455M Pre-A to Build Brain-First Embodied AI

Tashi Zhihang, a one-year-old Chinese embodied intelligence startup, has closed a $455 million Pre-A round, the largest single financing in China's embodied-AI history. The round was co-led by Hillhouse Capital and Sequoia Capital China, with continued participation from Meituan's strategic investment arm and multiple state-backed funds including the Beijing Robotics Industry Development Fund and Shanghai Guotou Pioneer Fund. About 20 investors across financial, strategic, industry, and state-owned categories participated, a composition the company says it deliberately curated.

Why it matters: This single round — larger than most Chinese embodied-AI companies' entire lifetime fundraising — signals that capital is concentrating on the 'full-stack brain' thesis: companies that vertically integrate data, hardware, and a unified foundation model for manipulation. The article explicitly frames the race as shifting from hardware flash (dancing robots, smooth locomotion) to functional brain capability (industrial dexterity, task generalization). Tashi Zhihang's choice of wire-harness assembly — an industrial 'impossible triangle' of long-horizon, flexible manipulation at sub-millimeter precision — as its proving ground is a deliberate bet that solving the hardest factory task unlocks every easier task. The company claims its AWE 3.0 world model, which unifies vision, language, and action in a native architecture, can perform failure recovery (re-plugging misaligned connectors) without any teleoperation data, instead using a human-centric capture paradigm called SenseHub.

Grounded expert take: This round validates the 'two-track' thesis: capital is now bifurcating between the brain-first full-stack players and the rest. The involvement of Meituan (a delivery-logistics giant with embodied-AI use cases in its warehouses) alongside top-tier VCs and provincial government funds is a textbook example of the 'strategic-plus-financial-plus-sovereign' syndicate that has become the hallmark of China's top-3 embodied-AI bets. The line that 'without a brain, the hardware body is just a shell' is the company's competitive positioning, but it also echoes the broader industry debate: does the full-stack brain moat hold, or will modularization split hardware and software? Tashi Zhihang is the clearest Chinese exponent of the former view.

#EmbodiedAI #Robotics #FullStackBrain #China #Funding #IndustrialAutomation

#Tashi Zhihang#embodied intelligence#full-stack brain#Pre-A funding#China AI#industrial robotics

How This Connects

Based on Robotics · Player Map

  1. 5d agoPrometheus raises $12B at $41B valuation, Bezos-led industrial AI startup.Prometheus
  2. 6d agoNeura Robotics raises $1.4B Series C for humanoid robots from Tether, Qualcomm, Amazon, NvidiaNeura Robotics
  3. 2w agoBYD unveils 4nm automotive AI chip Xuanji A3, claims 2100+ TOPS and dedicated NPU architecture
  4. 2w agoShift, an AI startup, has launched a free home cleaning service in New York City where workers wear...Shift
  5. 2mo agoTashi Zhihang (它石智航) Raises $455M Pre-A to Build Brain-First Embodied AI · THIS ARTICLE

Related News

More news from Aikido Security

Stay updated with the latest news and announcements from Aikido Security.

View all Aikido Security news

Discover AI Startups

Explore 2,000+ AI companies with VC-grade analysis, funding data, and investment insights.

Explore Dashboard