BYD confirmed to the South China Morning Post around July 30, 2026, that its first humanoid robot will debut in August at Di Space experience centers, and the company insists the machine on the floor will be a working interactive prototype rather than a fiberglass concept pose. The robot is called Xiao Di, also rendered Xiaodi in some coverage. It stands 1.61 meters, weighs 58.5 kilograms, and offers 31 degrees of freedom, with hands at 7 DOF each. The job description is showroom service: greet visitors, demo cars, and handle realtime translation across dialects and languages. Stella Li, BYD's executive vice president, has been framing a huge AI investment story and arguing China will be first to commercialize humanoids. Zhengzhou's Di Space already teased the arrival. The United States, meanwhile, got a July 28 reminder that foreign robots face a harder federal path after FCC Covered List moves.
A showroom robot with a real spec sheet
I care that BYD is leading with numbers instead of only a silhouette. 1.61 meters and 58.5 kilograms puts Xiao Di in a compact adult-ish envelope that can share aisles with shoppers without looking like warehouse hardware that escaped. Thirty-one DOF total, seven per hand, is not a research paper flex by itself, but it is enough detail to suggest the company expects people to ask what the hands can do while pointing at a car trim level.
The interactive prototype claim is doing important work. Automakers have burned years of goodwill with humanoid vapor that only walks in a branded film. BYD telling SCMP the August debut is a working unit raises the bar it now has to clear on camera. Greeting, demonstrating vehicles, and translating in realtime are measurable behaviors. Either the robot does them in a Di Space or the quote becomes a problem.
Showroom service is also a strategically modest first act. You do not need full factory autonomy to justify a greeter that can switch dialects for a family walking in from different regions of China. Translation is a BYD-flavored feature set because the company's domestic retail footprint already deals with language fragmentation as a customer-experience problem. A humanoid that fails at torque control in a plant can still win an afternoon if it nails a bilingual product spiel.
Zhengzhou's Di Space teaser matters as distribution geography. Experience centers are controlled environments with staff nearby, predictable flooring, and a narrative already about the future of the brand. That is the right cage match for a first humanoid. It is not the same as unsupervised night shifts next to stamping presses, and BYD is not pretending otherwise in the short-term plan.
I will be watching August for failure modes as much as for applause lines. Latency in translation, weird hand choreography near sheet metal, and the classic demo-handler hovering just out of frame will tell us whether this is product or stagecraft. Working prototype is a phrase companies use when they are proud and when they are hedging. Floor video will pick a side.
Stella Li's timeline, and the rumors BYD waved off
Stella Li's public line ties humanoids to a large AI investment thesis and to national first-mover ambition. China first to commercialize humanoids is the kind of sentence that plays in domestic press and reads like a dare everywhere else. I take it as strategic messaging. Commercialization, in her framing, clearly starts in branded retail space, not in a sudden consumer flood.
BYD's own horizon, as relayed in the coverage around the SCMP confirmation, is staged. Short term: two to three robots per store for sales assist. Medium term: malls and warehouses. Long term: homes. That ladder is almost suspiciously clean, which makes it useful as a corporate storyboard and dangerous as a literal calendar. Two to three units per store is a pilot density. Warehouses are a different reliability math. Homes are a liability novel.
Just as important is what BYD denied. Rumors had floated about 150 factory units, a Yao-Shun-Yu codename, and a 20,000-unit internal target. The company pushed those aside. I am glad it did so on the record, because humanoid discourse is drowning in leaked spreadsheet mythology. A denied rumor is not proof of the opposite. It is still a boundary around what we can responsibly repeat as fact.
Read together, the affirmation and the denials sketch a company that wants August credit for a showroom machine without inheriting every viral claim about factory armies. That is rational. It is also a tell that internal roadmaps and meme roadmaps have diverged enough to need a press correction.
If you are mapping competitive humanoid heat, put BYD on the automotive-retail track rather than the general-purpose android-from-another-planet track. The near-term KPI is whether Di Space visitors get a translation-assisted demo that feels better than a human salesperson with a tablet. Factory conquest can wait until the greeting works.
Why the U.S. path got harder in late July
On July 28, FCC action around a foreign robot Covered List complicated the United States path for equipment in this category. I am not going to turn a product story into a fake legal brief. The practical journalistic point is simpler. Even if Xiao Di dazzles in Zhengzhou, the export and deployment story into U.S. Venues is not a straight line from a good demo GIF folder.
That geopolitical friction cuts two ways. It may deepen BYD's incentive to win hard in China and other open markets first. It also means Western coverage should stop treating every Chinese humanoid debut as an imminent mall tour in California. The bottleneck is not only actuators. It is policy.
For BYD, August is still a brand moment inside a system it controls: Di Space centers, car shoppers, staff backup, and a corporate AI narrative Li is already selling. That is enough stakes. A working humanoid beside a flagship EV is a stronger commercial image than another concept reel, assuming the robot behaves.
My checklist for the debut is intentionally unromantic. Does Xiao Di greet without dead air. Do the 7 DOF hands demo a vehicle feature without a panic grab from a handler. Does realtime translation survive overlapping voices. Are there really two to three units planned per store, or does August produce one hero robot that tours. Answer those and you can ignore half the LinkedIn prophecy.
Humanoid hype cycles punish companies that over-promise factories and under-deliver handshakes. BYD, at least in the SCMP-facing message, is trying to promise the handshake. I will grade the August launch on that narrower exam. If the robot later graduates to malls and warehouses, fantastic. If homes show up in the same breath as a 1.61 meter sales assistant, I will reach for the same skepticism I use on every long-term slide. Show me the showroom first.
