The $34.5B Embodied AI Bet: Why VCs Are All-In on Robot Bodies

Let me start with a number: in the past 12 months, China’s embodied AI sector has absorbed over 345 billion yuan in funding.

To put that in perspective—that’s more than all AI investments combined in 2024. And this money isn’t spread thin. The top players are taking the lion’s share.

On April 16, Itashi Zhihang announced a $455 million Pre-A round, co-led by Hillhouse, Sequoia, and Meituan. Here’s the thing: this is a Pre-A round, not Series C or D. A company with essentially no revenue yet, valued at this level.

Honestly? I thought it was fake news when I first saw it.

What’s even more striking is that this round came just 8 months after their previous one. Investors were willing to keep pouring money in, at increasingly higher valuations, in rapid succession. What does that tell us?

It means they’re genuinely betting that the “mass production year” for embodied AI might finally be here.

Looking at the timeline: 2023 was the concept explosion year—demos everywhere. 2024 was about finding real use cases. 2025 saw valuation madness. And 2026 seems to be shaping up as the “mass production validation year.”

Why 2026? Two reasons.

First, the April 19 humanoid robot half-marathon in Beijing’s Yizhuang was a landmark. Over 100 teams ran 21 kilometers. This wasn’t a stunt—it was a real capability test. Glory’s “Qitian Dasheng” took the championship, finishing in 50 minutes and 26 seconds. You can say we’re still far from practical application, but you can’t deny we’ve jumped multiple orders of magnitude in capability compared to two years ago.

Second, the supply chain is maturing. The core bottleneck for embodied AI mass production used to be component costs—joint motors, reducers, sensors. Now domestic suppliers have solutions that can push整机成本 to under 200,000 yuan per unit. Still expensive, but 80% cheaper than two years ago.

So the logic behind this funding boom is: technology has crossed the usability threshold, the supply chain has crossed the cost threshold, and now it’s about who can ship volume first.

But here’s my question: among all these billions, how much is actually pushing technical boundaries versus just chasing deals?

I won’t jump to conclusions, but I know one thing: the hotter the track, the harder you need to think clearly about individual projects.

Next checkpoint is Q3 this year. Let’s see how these well-funded players progress on actual mass production.

We’ll revisit this then.