XPeng Iron Price 2026: Is It For Sale, Cost & Release Date

Is the XPeng Iron for sale, and what does it cost?

No, you cannot buy an XPeng Iron in July 2026, at any price, from anyone, and XPeng has never announced one. What XPeng has committed to is mass production by the end of 2026, which is not the same thing as a sale date: the first Irons are slated to work in XPeng's own Chinese showrooms from early 2027, go overseas later that year, and reach households in 2028 at the earliest. Every dollar figure you've seen attached to Iron is third-party guesswork, one "listing" selling it is fake, and there's a cryptocurrency squatting on its name. We verified all of it against primary sources on July 22, 2026.

The is-it-real saga, told precisely

Iron is famous because people refused to believe it. On November 5, 2025, at XPeng's AI Day in Guangzhou, the next-generation Iron walked a catwalk with a gait smooth enough that Chinese and Western social media concluded a human was inside the suit, citing an "ear" outline (actually a microphone array) and contours under the back covering. XPeng's response came the next morning, November 6, not "days later" as widely retold: He Xiaopeng posted a single-take, unedited video in which a technician unzips the robot's covering to show a mechanical torso, and later that same day, on stage at a separate launch event, staff cut open Iron's leg with scissors while it stood powered on, exposing actuators and wiring. Elon Musk liked the video. The hashtag for what Iron looks like without skin topped Douyin.

Two honest footnotes the viral clips skip. First, robotics analyst Scott Walter argued from frame analysis that the robot walking on stage was the older-generation chassis running new gait software, with the next-gen hardware appearing only in renders; XPeng never directly rebutted him. Second, Iron's first public outing after the reveal, at a Shenzhen mall on February 1, 2026, ended with the robot falling backwards after its catwalk; it appeared the next day strapped to a support frame, and He Xiaopeng compared it on Weibo to a toddler learning to walk. The robot is real. It is also still learning to stand around unpredictable humans.

The price truth: three debunks in one section

XPeng has never stated a price. He Xiaopeng's only on-record framing is directional, and it's about cost, not price: the manufacturing cost of humanoid robots could eventually approach that of cars. That is not a retail-price announcement. From there, three things pollute the search results. The "$150,000" figure repeated across robot-listing sites is an analyst-style estimate none of them source to XPeng. A marketplace called robotixmarket.com displays a "$120,000 — In Stock" Iron listing with an add-to-cart button for a robot that no one, anywhere, can purchase; treat it as decoration at best. And an unrelated "XPENG IRON" meme token trades on BNB Chain and Solana, which is why crypto price pages rank when you search the robot's cost. None of these is XPeng, and none of them knows the price, because there isn't one.

Specs, by generation (this is where everyone gets it wrong)

Most competitor pages mix two different robots. The 2024 Iron and the 2025 next-gen Iron share a name and little else; the figures below are XPeng's own claims, labeled, with the conflicts shown.

The aggregator numbers you'll see elsewhere — "4-hour runtime," "7 km/h," "20 kg grip" — appear on spec farms only and trace to no XPeng statement; we've omitted them.

What Iron actually does today

Here the honest record diverges sharply from the 2024 marketing. XPeng said then that Iron was "already operational" building EVs in its Guangzhou factory. A year later, 36kr's reporting found the factory fleet was used mostly for algorithm tuning and teleoperation data collection, and He Xiaopeng himself admitted the screw-tightening experiment failed economically: the hands wore out within a month and cost more to replace than years of a worker's wages. That admission drove the strategy you see now: commercial service first (showroom guide, receptionist, tour host), selective industrial inspection with sole named partner Baosteel, and households last. It is a more credible plan precisely because it was arrived at the hard way.

Iron's 2026, dated

Worth noting what's absent: Iron skipped July's WAIC in Shanghai, where rivals AgiBot and Unitree filled the halls; XPeng ran its own Munich show instead. And the leadership churn behind the June takeover cuts both ways: it's the strongest possible commitment signal, and it's also three senior departures in the year before the hardest deadline the company has ever set.

The female-design question, using XPeng's actual words

The next-gen Iron's feminine styling generated a commentary cycle largely detached from what XPeng said. The company's stated position: the design is customizable, with female and male body variants planned, because buyers will choose body types the way they choose car trims, and service robots should feel approachable. The robotics VP put it flatly: "We see the different options because there are different kinds of humans." Whether the launch-choreography choice to debut the female variant on a catwalk was wise marketing is a fair debate; the claim that XPeng builds only a "female robot" is simply wrong.

Iron vs Optimus: the only comparison that matters

The honest framing: Iron and Optimus are two of the most closely watched unpurchasable robots on Earth, both promising the same year, both from EV makers betting the factory. Iron's differentiators are anthropomorphism (the skin, the gait, the body variants) and a nearer commercial venue (its maker's own showrooms). Optimus's are scale ambition and the Tesla brand. Neither has shipped a product. Our evidence-ranked guide covers the robots that actually have.

The stock question

There is no XPeng robotics IPO. Searchers conflate Iron with XPeng's listed EV stock (XPEV) and with rival Unitree's approved listing; the actual July 2026 signal points the other way, with He Xiaopeng consolidating the robotics unit under his personal control rather than spinning it out. XPEV's valuation debate now partly rides on Iron: at least one analysis argues the stock is substantially overvalued on the robot push. If you want listed humanoid exposure with actual shipments behind it, that conversation currently runs through Unitree's pending IPO, covered in our G1 review.

Misinformation audit: eight Iron claims, checked

What to watch next

Iron's calendar is unusually checkable. End of 2026: does mass production actually start, at anything like 1,000 units a month? Q1 2027: do Irons appear in XPeng showrooms doing real shifts? And between now and then, whether the promised European availability gets a date and whether XPeng publishes the number everyone is searching for. Every one of those has a falsifiable deadline, which is more than most humanoid roadmaps offer, and we'll update this page as each lands or slips.

The bottom line

The XPeng Iron is real, remarkable, and not for sale. It owns one of the most humanlike publicly demonstrated gaits in the industry and the scars to prove it's a machine; it also owns a public fall, an unrebutted question about which chassis actually walked that catwalk, and a factory story its own CEO admits didn't work the first way. If the end-2026 production line and the Q1 2027 showroom jobs happen on schedule, Iron would give ordinary showroom visitors routine access to one of the most humanlike service robots yet deployed. Until then it is the best advertisement yet for a product that does not exist, and every price you read for it, including any future one here, deserves the same test: did XPeng say it, or did someone guess?