Embodied AI Fuels Record Robotics Funding In China As IPO Momentum Builds
Summary
China's robotics sector has achieved a record-high venture investment, reaching \$5.6 billion across 176 deals by mid-May 2026, matching the total for all of 2021 and surpassing the \$4.3 billion raised in 2025. This surge is driven by a shift towards embodied AI, where startups integrate artificial intelligence with physical bodies for real-world interaction, moving from pre-programmed hardware to Vision-Language-Action models. Notable companies like TARS Robotics (\$513 million seed round, \$1.9 billion valuation), X Square (\$140 million Series A extension, \$293 million Series B), Spirit AI (\$435 million Series A total, \$1.5 billion valuation), Galaxea AI (\$435 million Series B total, \$1.4 billion valuation), and EngineAI (\$200 million Series B, \$1.5 billion valuation) have secured significant funding. IPO momentum is building, with Unitree Robotics targeting a \$3-7 billion valuation and Robotphoenix listing on HKEX, raising \$86 million and closing up nearly 80%. Additionally, AgiBot acquired a 63.62% stake in Swancor Advanced Materials for \$290 million in July 2025.
Key takeaway
For investors evaluating the global AI landscape, China's robotics sector presents a significant opportunity, driven by the rapid adoption of embodied AI and strong IPO momentum. You should consider allocating capital to startups focusing on software-hardware integration and Vision-Language-Action models, as these are attracting substantial funding and achieving high valuations. Monitor the Shanghai and Hong Kong exchanges for upcoming robotics IPOs, as these exits signal market maturity and potential for future returns.
Key insights
China's robotics funding surge is fueled by embodied AI, shifting investment from hardware to integrated AI-driven systems.
Principles
- Embodied AI drives robotics innovation.
- Vision-Language-Action models enhance robot autonomy.
- Software-hardware integration attracts large investments.
Method
Robotics development is transitioning from line-by-line coding to using advanced open-source reasoning models, enabling physical machines to observe, reason, and execute tasks end-to-end.
In practice
- Investigate Vision-Language-Action models.
- Focus on software-hardware integration.
- Explore Hong Kong for robotics IPOs.
Topics
- Embodied AI
- Robotics Investment
- Humanoid Robotics
- Venture Capital China
- IPOs
- Vision-Language-Action Models
Best for: Investor, Entrepreneur, Director of AI/ML
Related on AIssential
Editorial summary, takeaway, and curation by AIssential. Original article published by Artificial intelligence - Crunchbase News.