Funding & Businessroboticsfundingembodied aiphysical ai

X Square Robot Tops $2.8B Valuation for Embodied AI

||By LDS Team
6.5
Relevance Score
X Square Robot Tops $2.8B Valuation for Embodied AI

For teams tracking physical AI, the signal worth noting is how fast Chinese capital is consolidating behind embodied foundation models rather than narrow task robots. X Square Robot has closed four consecutive financing rounds ending in a Series C, pushing its valuation above $2.8 billion (about 20 billion yuan), the company said on June 29, 2026. IDG participated in the Series C, while HongShan and Xiaomi backed earlier rounds; combined with lead investments from Alibaba, ByteDance, Meituan, and Xiaomi, the startup says it is the only Chinese embodied AI company to win lead-round backing at different stages from all four of the country's top technology firms. X Square is building a full-stack embodied AI system spanning foundation models, robotics hardware, and a data pipeline, anchored by WALL-B, a model introduced in April 2026 that trains perception, language, action, and physical prediction in one network.

Why it matters

The architecture claim is the most interesting part of this raise. X Square says its WALL-B model, introduced in April 2026, trains perception, language, action, and physical prediction inside a single World Unified Model network rather than bolting separate vision-language-action components together. If that approach holds up, it points toward embodied systems that reason about physics and learn continually from real-world interaction, which is the harder problem behind general-purpose robots.

What was announced

X Square Robot disclosed four consecutive financing rounds culminating in a Series C, lifting its valuation above $2.8 billion (roughly 20 billion yuan), according to a company announcement dated June 29, 2026. IDG participated in the Series C, while HongShan and Xiaomi backed prior rounds. The company says lead investments across its history came from Alibaba, ByteDance, Meituan, and Xiaomi, making it, by its own account, the only Chinese embodied AI startup to secure lead-round backing at different stages from all four firms.

Practitioner read

The funding concentration matters as much as the dollar figure. When four of China's largest technology companies anchor a single embodied AI startup, it concentrates compute, data, and deployment channels in a way that can accelerate iteration on world-model robotics. The hard questions stay empirical: how WALL-B performs on real manipulation and navigation tasks, how much of its capability transfers across robot bodies and environments, and whether the unified-network approach scales better than modular pipelines.

What to watch

  • Independent evaluations or demonstrations of WALL-B on manipulation and long-horizon tasks.
  • How the company sources and curates real-world interaction data, the binding constraint for embodied models.
  • Whether US-China competition in physical AI draws the export and policy scrutiny already seen in chips and frontier models.

Key Points

  • 1X Square Robot closed four consecutive rounds ending in a Series C, pushing its valuation above $2.8 billion for embodied AI.
  • 2All four leading Chinese tech firms, Alibaba, ByteDance, Meituan, and Xiaomi, have led rounds, concentrating domestic capital behind physical AI foundation models.
  • 3Heavy funding for unified world-model robotics shows China racing to commercialize general-purpose embodied systems amid intensifying competition with the United States.

Scoring Rationale

A Series C lifting an embodied AI startup above a $2.8 billion valuation, backed by all four leading Chinese tech firms, is a notable signal of momentum in physical AI foundation models. It matters to practitioners tracking world-model robotics and the geopolitics of embodied AI. It falls short of major status absent independent benchmarks or a single round exceeding $500 million.

Practice interview problems based on real data

1,625 SQL & Python problems across 15 industry datasets — the exact type of data you work with.

Try 250 free problems