BYD's first humanoid robot design just surfaced

BYD's first humanoid robot design just surfaced

John KoetsierOctober 10, 2026
ShareXLinkedInFacebook
BYD filed a design patent for its humanoid robot this week, giving us the first official look. Behind it: 4,000 engineers, a seventh-generation prototype, a self-developed 4nm chip, and a target of 20,000 internal factory deployments by year-end.

BYD didn't announce a humanoid robot concept this week. Instead, it confirmed a program, named a codename (Yao Shun Yu), revealed that it has already deployed 2,000 prototype units last year, and set an aggressive target of 20,000 internal factory deployments by the end of this year. 

Reportedly, BYD has 4,000 engineers on the project.

We got our first look at the hardware as China's National Intellectual Property Administration published a design patent that BYD filed for an "embodied intelligent robot," giving us the first official visual of the machine. Bipedal, roughly human proportions, oval-shaped head, articulated hands, exposed mechanical structures at the joints.

Image

Of course, this is just a rough image from a patent. Expect the actual robot to look significantly different.

The numbers behind the stealth build

According to reports, BYD's seventh-generation prototype walks at 1.5 meters per second, can carry 50 kilograms, and has a four-finger hand with 16 degrees of freedom. It runs a LiDAR plus 6D tactile sensing system for perception.


By the company's own benchmarking, it achieves roughly 80% of human labor efficiency on factory tasks. 

Interestingly, over 60% of the robot's core components share specifications with BYD's automotive supply chain. The company will manufacture 80% of its parts in-house, and servo motors cost will apparently cost 40% less than external suppliers charge. Proprietary reducers come in at 60% below imported equivalents.

The Xuanji A3 chip for BYD's humanoid robot

On May 28, 2026, BYD unveiled the Xuanji A3, a self-developed 4nm chip. 

The company's stated use case covers both automotive AI and robotics. This matters because supply chain vulnerability in advanced semiconductors is one of the real constraints on Chinese humanoid makers scaling up. BYD is trying to close that loop with in-house silicon.

Most humanoid startups buy chips from NVIDIA or Qualcomm. BYD is building its own. We'll see if that was a smart move when we see what the robot is capable of live.

Why the August debut never happened

Li Ke, BYD's executive vice president, confirmed the program publicly on June 3, 2026. By late July, BYD confirmed plans to show the robot at its Di Space showroom format in August. 

October, of course, has arrived, but  the public debut has not.

This isn't unusual for a company of BYD's profile. In EVs, BYD has repeatedly preferred to announce products when they're ready to sell rather than stage early reveals. The prototype units currently running at BYD facilities in Shenzhen and Changsha suggest they're not stalling for engineering reasons, and the October design patent publication might be a sign that an official launch is coming soon.

The open platform bet

Executive vice president Stella Li has described BYD's goal as building an open robotics platform: one that accommodates both BYD's own robots and products developed in partnership with other companies. She also envisions 2-3 robots deployed at every BYD dealership worldwide to assist with customer interactions.

The open platform framing is the most interesting strategic signal in the entire program. Every other serious humanoid manufacturer is building a closed ecosystem. China's humanoid market has dozens of competing startups, all trying to own the full stack. BYD is large enough that it could offer something different: an infrastructure layer that others build on top of.

Whether that plays out depends on execution. BYD has tried open-ish platform strategies in EVs with mixed results, so we'll see.

What to watch

The formal product debut is the obvious near-term signal. 

If it happens in October or November, BYD will have slipped its August deadline by two to three months, which is a rounding error for a program that has been running for four-plus years. If the debut slips into 2027, that changes the picture considerably.

The 20,000 internal deployment target for 2026 is probably at risk, given the calendar right now. BYD wanted 20,000 in its own factories by year-end. That's an enormous target, and one that even BYD with its massive manufacturing scale is likely not going to be able to hit.

Li Ke said the competition comes down to "manufacturing capability, software capability, and hardware capability." On the first dimension, few other companies can match BYD. 

Whether the software and AI are great, however, is another question.

Sources

Share this article

ShareXLinkedInFacebook

Written by

John Koetsier
John Koetsier

I tell stories with data and build startups. I've raised angel and VC funds, led software companies and startups, and chronicled the rise of the mobile economy. Currently I write for Forbes, host several podcasts, and am starting up HumanoidMegaHub. I was TUNE's mobile economist and Singular's VP of Insights, built the research team at VentureBeat, and managed teams creating software for partners like Intel and Disney. In past lives I built websites, e-learning platforms, and mobile apps, and, a long time ago, was named to Folio's top 100 of the media industry's "most innovative entrepreneurs and market shaker-uppers.”