China Robotics Funding Landscape: H1 2026 — Who's Raising, Who's Exiting
Executive Summary
China’s robotics funding market in H1 2026 presents a study in extreme contrasts. Blue-chip startups raised massive rounds — Leju Robotics $360M, Galbot $100M+, Zhiyuan $70M — while others like Kepler sold for less than 30% of peak valuation. Behind this polarization lies a market returning to rationality: strong companies get strong capital, weak companies get pushed out. Meanwhile, Unitree’s IPO filing and XPeng’s announced mass production of humanoid robots signal that the industry is moving from PowerPoint fundraising to delivery validation.
Funding Overview: H1 2026 Major Events
| Company | Round | Amount | Investor/Acquirer | Notes |
|---|---|---|---|---|
| Leju Robotics | Strategic financing | ~$360M (26B RMB) | Undisclosed | One of largest single robotics raises in China |
| Galbot | Series A | $100M+ | Multiple VCs | Full-stack humanoid startup |
| Zhiyuan Robotics | Series A | $70M | Lei Jun’s investment network | Xiaomi ecosystem-linked |
| XinZhi Embodied Intelligence | Angel | ~$14M | Undisclosed | Embodied intelligence newcomer |
| Kepler (exit) | Acquired | ~$40M (3B RMB) | Hangzhou Kolin | 70%+ discount from peak valuation |
| Unitree Robotics | Pre-IPO | — | — | Filing for IPO |
| XPeng Motors (internal) | — | — | Self-funded | Humanoid robot mass production by year-end |
Sources: Public funding announcements, 36Kr, Jiqizhixin (Machine Intelligence) compilation. Some transaction amounts are market estimates.
Player Breakdown
Hardware Companies: Products Speak
Unitree Robotics was the biggest winner of H1 2026. The G1 humanoid robot ships globally at $16,000 — controversial in capability (motion control and software ecosystem lag behind Tesla Optimus significantly), but “shipping first” is a victory in itself. Unitree is now preparing for IPO. As one investor told 36Kr: “If Unitree can deliver a revenue report above RMB 100M by 2027, the IPO won’t be difficult.”
Leju Robotics raised approximately $360M — one of the largest single raises in the humanoid sector this year. Leju’s strategy differs from Unitree: more focus on industrial deployment than consumer demonstrations. For context, $360M is roughly one-third of Figure AI’s Series B size. Given China’s labor and manufacturing cost advantages, Leju’s $360M may take them further than Figure AI’s $675M.
AI Software Companies: The Rise of Embodied Intelligence
Galbot and XinZhi Embodied Intelligence represent the surging investor interest in “embodied intelligence.” Galbot’s $100M+ Series A is the second-largest humanoid raise this year. Founded by former iFlytek researchers, the company takes a full-stack approach — from actuators to perception algorithms, all developed in-house.
Zhiyuan Robotics raised $70M in Series A, led by capital tied to Xiaomi founder Lei Jun. As part of the Xiaomi ecosystem, Zhiyuan’s advantage lies in consumer electronics channel access and supply chain resources. If Xiaomi launches a humanoid robot in the future, Zhiyuan would be the technology core.
M&A Activity: Market Consolidation Begins
Kepler Robotics was acquired by Hangzhou Kolin for approximately $40M — less than 30% of its peak valuation. This case should serve as a warning to every robotics startup.
Kepler was valued above $140M in 2024 on the strength of an early humanoid demo that attracted significant attention. Two years later, the product failed to reach mass production, technology direction shifted repeatedly, and investors lost patience. The exit was effectively a distressed sale — part acquisition, part bankruptcy restructuring.
The signal is clear: the market no longer rewards “demo companies.” The gap from demo to mass production has become the dividing line between survival and failure.
Key Trends
Trend 1: Valuation Divergence Intensifies
Leju at $360M and Kepler at $40M sit at opposite extremes of the same sector. This is no accident.
In 2024-2025, the humanoid robot concept was white-hot; almost any team with a working demo could raise capital. 2026 tells a different story. Capital has become more disciplined, using “production capability” and “revenue data” as valuation benchmarks.
China’s robotics valuation framework is shifting from “compare to overseas companies” to “read the Chinese financials.” A partner at a top-tier domestic VC put it bluntly: “We no longer benchmark against Figure AI or Tesla Optimus. We look at where this company’s factory is, whether the production line is running, and where the customer orders are.”
Trend 2: From “Series C Death” to IPO Wave
Unitree’s IPO filing is a milestone for China’s robotics industry. Previously, the typical fate for Chinese robotics startups was either acquisition by a tech giant or “Series C death” — burning through capital with no path to the next round.
Whether Unitree’s IPO succeeds at a reasonable valuation will define the ceiling for the entire sector. If it does, it opens exit pathways for Leju, Galbot, and others. If the stock trades below its IPO price, it could trigger a valuation correction across the board.
Trend 3: Big Tech Enters, Accelerating Polarization
XPeng (XPeng Motors) announced that its humanoid robot will reach mass production by the end of 2026 — a timeline ahead of most pure robotics startups. XPeng’s advantage is clear: automobile manufacturing is robotics’ closest industrial relative, sharing precision manufacturing requirements and substantial supply chain overlap.
Xiaomi’s CyberDog 2 robot dog is already on sale through Xiaomi’s online store. While not humanoid, CyberDog has accumulated experience in perception, motion control, and mass production — capabilities that can transfer to a humanoid product when the time comes.
For startups, the competitive landscape has shifted from “competing with other startups for funding” to “competing with Xiaomi and XPeng for talent, supply chain access, and market share.”
Trend 4: M&A Accelerates
Kepler will not be the last robotics company acquired.
The dozens of robotics startups that emerged in 2024-2025 face a common fate: product fails to ship, funding dries up, team dissolves. The lucky few will find acquirers at distressed prices — at least the founding team exits with dignity.
Expect more robotics M&A deals in 2027. Acquirers will include both listed companies (buying technology and teams) and overseas companies looking for a channel into China’s robotics market.
Investor Perspective: Which Tracks Are Attracting Capital?
Ranked by current funding heat:
- Full-stack humanoid robots (Leju, Galbot, Zhiyuan) — largest funding rounds, but most competitive
- Embodied intelligence software (XinZhi, others) — rising quickly in investor attention, commercialization path still unclear
- Robot core components (actuators, gearboxes, torque sensors) — B2B model with healthier cash flow, attracts industrial capital
- Scenario solutions (inspection, logistics, warehouse robots) — smaller raises, but higher certainty of deployment
The strongest investor preference is for “humanoid robot companies with mass production capability.” This means:
- Pure software teams without hardware capability will struggle to raise
- Pure hardware teams without AI software expertise will struggle to command premium valuations
- The valuation gap between “mass production ready” and “lab prototype” is at least 10×
Risk Factors
- Valuation bubble burst: If Unitree’s IPO disappoints, the entire sector’s fundraising environment could deteriorate sharply
- Production hell: Humanoid robot mass production is far harder than capital markets anticipate. Even Tesla Optimus has not achieved genuine volume delivery as of mid-2026
- Technology route uncertainty: Bipedal locomotion combined with dexterous manipulation has not converged on a dominant design. Companies betting on wrong technical approaches will be eliminated
- Policy risk: Robotics intersects with AI and security regulation. US-China trade friction may disrupt core component supply
Independent Assessment
The biggest signal from H1 2026 funding data is this: the humanoid robotics industry is transitioning from storytelling to delivery.
Leju’s $360M raise is not a sector-wide phenomenon — it represents a capital feast available only to the top tier. For the vast majority of robotics startups, 2026-2027 means either achieving mass production, finding a buyer, or shutting down. The gray area is disappearing.
Bet on companies that can ship. Be wary of companies with great slide decks. We expect H2 2026 funding data to show even more extreme polarization.
Last updated: June 1, 2026