Xpeng's humanoid robot: The race for physical AI kicks off
MuckyNews reading
Xpeng, primarily known as an electric vehicle maker from Guangzhou, is making a serious push: by the end of 2026, they plan to produce over 1,000 humanoid robots per month, with global commercial deliveries of their model "Iron" projected for 2027. The goal is bold – CEO He Xiaopeng says they want to sell a million of these machines by 2030. The promise is that these robots will become affordable for households, which really just means one thing: Xpeng wants mass production and aggressively lower costs.
What’s happening here is a pivot from carmaker to a player in so-called physical AI – the integration of intelligent software into physical machinery. Xpeng relies on an in-house AI model that supposedly converts visual data directly into motion, skipping text-based code. That sounds like a technical edge, but in marketing terms it mainly serves one purpose: creating distance from Tesla.
Because that’s the real context. Tesla's Elon Musk has already repurposed his Fremont factory to build the Optimus humanoid, though it hasn't been commercially sold yet. That’s exactly where Xpeng steps in, positioning itself not as a copycat, but as the one who actually delivers first. The choreography is obvious: we’re not imitators, we’re the first to market. Whether production can actually match these announcements, and whether a million robots by 2030 is anything more than a bold projection, remains a completely different story.
Short version
- Xpeng plans mass production of its humanoid robot "Iron" by late 2026,
with global sales starting in 2027.
- The company aims to sell one million robots by 2030.
- The technology relies on directly converting visual data into physical
motion.
- The move is framed as a direct challenge to Tesla's physical AI
ambitions.
- The real pressure point: whether Xpeng can actually scale production and
sales as fast as they announce them.
Perspective delta
hongkong_china_asia: The South China Morning Post coverage clearly frames this as a head-to-head race against Tesla. Xpeng is built up as the proactive challenger, beating Tesla to the punch because Optimus isn't commercially available yet. The emphasis is on technological independence and Chinese innovation. The massive gap between currently producing zero units and projecting a million sales by 2030 gets barely any scrutiny. The economic and technical risks of scaling this fast are largely left out.
Source base / limits
The source base is very narrow. There is only a single article from the South China Morning Post. This means the entire picture is shaped by this one outlet and the corporate statements provided by Xpeng. There are no independent expert sources beyond a Shanghai consultant quoted in the same piece. The production forecasts and sales targets cannot be verified through other sources.
Sources / basis
- MuckyNews container ID: MN-20260716-082
- Source outlets present: South China Morning Post (SCM)
- Perspective groups present: hongkong_china_asia
- Article count: 1
- Source count: 1
- Perspective count: 1
- Approximate raw text size: 2085 characters