Industry News
Xiaomi and Sequoia Back $276M Bet on Home Humanoid Robots
X Square Robot, a Shenzhen-based embodied AI startup, raised nearly $276 million in a Series B co-led by Xiaomi and Sequoia China, with participation from ByteDance, Meituan, and Alibaba. The company is targeting paid home trials through partner 58.com by late May 2026, making it one of the most aggressive household humanoid timelines globally. The round brings X Square's total funding to approximately $580 million.
X Square Robot, a Shenzhen-based embodied AI startup, raised nearly 2 billion RMB (approximately $276 million) in a Series B round announced April 22, 2026 -- one of the largest single funding events in China's household robotics sector this year. The round was co-led by Xiaomi's strategic investment arm and Sequoia China (HongShan), with participation from ByteDance, Meituan, and Alibaba.
The company builds general-purpose humanoid robots designed for home use, powered by its WALL-B and WALL-OSS foundation models. With the new capital, X Square plans to launch paid home trials by late May 2026 -- the first commercial home deployment test for a household humanoid in China at this scale. That timeline is four years ahead of where most Western competitors say home robots will be viable.
AI-generated image
Home environments present complex, unstructured challenges for humanoid AI systems. Source: Biped.news
Key Stats
$276M
Series B Raised
May 2026
Home Trial Launch
4
Lead Investors
~$580M
Total Funding Raised
The Household Humanoid Bet
Most of the humanoid robotics money flowing through 2025 and early 2026 has gone toward factory and warehouse applications. The logic is straightforward: industrial environments are structured, tasks repeat, and the economics of replacing one $35-per-hour logistics worker with a $70,000 robot amortized over three years are easy to model. Homes are a different problem entirely -- unstructured, full of fragile objects, occupied by unpredictable people and pets, and with no clear throughput metric to optimize.
X Square Robot is betting that the home problem is more tractable than it looks, and today's funding suggests Xiaomi and Sequoia China agree. The company was founded by engineers from leading Chinese AI labs and has been working specifically on the home use case since its earliest days, rather than pivoting from industrial robotics. That focus shows in its model architecture: WALL-OSS and WALL-B are trained on household task data, not factory manipulation data.
The $276 million round builds on a $100 million Series A in 2025 and a $140 million A++ round in early 2026, bringing total funding to approximately $580 million. X Square has not disclosed its valuation, but comparable Chinese embodied AI companies at similar stages have been valued between $1.5 billion and $3 billion.
Home-First Architecture
X Square's WALL-B model was trained on household episode data from the start, not adapted from industrial manipulation models. The system handles dynamic, unstructured environments by design rather than as an afterthought.
58.com Home Deployment
X Square plans to deploy early units through partner 58.com, China's largest classifieds and services platform. That channel gives the company direct access to household customers for the May 2026 trial rollout without building its own sales infrastructure.
WALL-B Foundation Model
The WALL-B model handles task planning and execution for cleaning, cooking assistance, and elder care tasks. It chains sub-tasks using a hierarchical planner that breaks complex instructions into discrete manipulation primitives.
Multi-Hour Operation Target
X Square has not published formal battery life specs, but the company describes household task completion across 6-8 hour windows as a design target, consistent with the working patterns of a domestic helper.
Under the Hood: Embodied AI for the Home
The technical challenge separating household humanoids from their industrial counterparts comes down to one thing: generalization. A factory robot can be trained on a few thousand examples of a specific task in a controlled workspace and perform adequately. A home robot encounters a different kitchen, different drawer handles, different lighting conditions, and a different family's habits in every deployment. The training data required to cover that variance is orders of magnitude larger than what industrial deployments require.
X Square's approach to this problem centers on its WALL-OSS (open-source) and WALL-B (behavioral) model stack. WALL-OSS handles perception and scene understanding -- mapping the environment, tracking object state, identifying surfaces and containers. WALL-B handles action planning and physical execution, translating a task like "clear the breakfast table" into a sequence of grasps, placements, and navigation waypoints. The two models communicate at approximately 10 Hz, with WALL-B querying WALL-OSS for updated scene state after each manipulation step.
AI-generated image
The WALL-B and WALL-OSS model stack coordinates perception, planning, and physical execution. Source: Biped.news
The hardware platform has not been publicly fully specced, but X Square has shared some configuration details: the robot is human-scale (approximately 165-170 cm), uses electric quasi-direct-drive actuators in the arms and legs, and carries onboard compute capable of running both models without cloud dependency. The vision system relies on multi-camera RGB-D fusion, similar in principle to the sensor stacks used by Figure AI and 1X Technologies.
Household AI Robot Comparison: X Square vs. Peers
| Company | X Square Robot | Sunday (Memo) | 1X Technologies (NEO) | UniX AI (Panther) |
|---|---|---|---|---|
| Target Market | Home (China) | Home (US) | Home / Light commercial | Home (China) |
| Total Funding | ~$580M | ~$230M | ~$100M | Undisclosed |
| Foundation Model | WALL-B / WALL-OSS | Proprietary (10M+ episodes) | OpenAI-trained | Proprietary |
| Form Factor | Bipedal humanoid | Wheeled base + arms | Bipedal humanoid | Bipedal humanoid |
| Home Trial Status | May 2026 (planned) | Thanksgiving 2026 (planned) | Beta program 2026 | Deploying (Suzhou, April 2026) |
| Lead Investors | Xiaomi, Sequoia China | Coatue, Sequoia | Andreessen Horowitz, OpenAI | Undisclosed |
Training Data Strategy
- Simulated data: X Square uses physics simulation to generate large volumes of household task demonstrations across varied environments. This covers edge cases that would be expensive to collect in the real world.
- Real episode collection: The company employs internal operators to run the robot through household tasks in instrumented test apartments, capturing high-quality manipulation data.
- 58.com integration: The planned home trial through 58.com will generate in-the-wild deployment data at scale -- the kind of distribution coverage that simulations cannot fully replicate.
Who Backed This Round and Why It Matters
The investor lineup in X Square's Series B tells a story about where China's tech giants think the robot industry is going. Each of the five major participants brings something beyond capital.
- Xiaomi (co-lead): China's smartphone and consumer electronics giant has its own CyberOne humanoid in development and began factory trials at its EV facility in March 2026. Backing X Square is not a contradiction -- it's a hedge. If the household humanoid market grows large enough, Xiaomi wants equity in the winner regardless of whether that's its own product. The strategic overlap also raises the possibility of hardware or ecosystem integration down the line.
- Sequoia China / HongShan (co-lead): HongShan has backed several of China's most successful hardware companies. The firm co-led X Square's Series A and doubled down with a co-lead position in the B, which signals internal conviction that the company is on track against plan. HongShan rarely leads back-to-back rounds in robotics without hard evidence of technical progress.
- ByteDance: TikTok's parent has been quietly building robotics research capabilities and has made a handful of strategic investments in the space. Its participation here suggests an interest in the data and content layer -- home robots generate enormous amounts of behavioral and household data that could be relevant to ByteDance's AI ambitions.
- Meituan: China's dominant food delivery and local services platform. If X Square's robots can eventually handle meal preparation or delivery coordination inside the home, Meituan's ecosystem is an obvious integration point. The investment is both financial and strategic -- Meituan gets early visibility into a technology that could reshape its last-mile service model.
- Alibaba: Alibaba has been investing broadly in physical AI and robotics infrastructure. Its participation in this round adds an e-commerce and logistics angle -- Alibaba operates some of the world's largest smart warehouses, and household AI systems trained on domestic environments could eventually extend into last-mile service scenarios.
Why the Investor Mix Matters
Every company in this round has a direct strategic interest in household robotics, not just financial returns. That's different from a venture capital round with generic tech investors. It means X Square has relationships with potential distribution partners (58.com, Meituan), hardware ecosystem partners (Xiaomi), and cloud infrastructure (Alibaba) built into its cap table. That network could compress the go-to-market timeline significantly.
What This Means for the Household Robotics Race
The household humanoid market has been treated as a 2030+ proposition by most Western analysts. The argument has been that factory and warehouse applications were the necessary first step -- building technical capability and revenue before taking on the harder home problem. X Square's timeline disrupts that sequencing. If home trials launch in May 2026 and generate usable data, a commercial product could arrive in Chinese homes by 2027 or 2028, well ahead of what companies like Sunday Robotics (targeting Thanksgiving 2026 in the US, for a wheeled non-humanoid product) are projecting.
The China-specific context matters here. Labor costs in China's major cities have risen steadily, aging population trends are acute, and there's no equivalent cultural resistance to domestic robots that exists in some Western markets. The Chinese government has also listed humanoid robotics as a strategic industry, which means regulatory approval processes for home deployment trials may move faster than they would in the US or EU.
For Western companies in this space, the X Square round is a signal worth watching. A $580 million total raise, Xiaomi and Sequoia behind it, and home trials six weeks out -- that's not a research project. That's a company building toward product-market fit. The gap between Chinese household humanoid timelines and Western ones appears to be widening, not closing.
~$580M
X Square Total Funding
May 2026
Target Home Trial Date
5
Strategic Corporate Investors
What's Coming Next
The immediate test is May 2026. X Square has committed to launching home trials through 58.com within six weeks of this funding announcement. The success criteria for that trial are not public, but the company will almost certainly be tracking task completion rates across a defined set of household chores, failure modes by environment type, and customer satisfaction scores. That data will determine whether a broader commercial rollout follows in late 2026 or early 2027.
Watch for a few things in the next six months: whether X Square publishes trial results or demo footage from real homes (as opposed to instrumented lab environments), whether Xiaomi moves to formally integrate the company's technology into its broader smart home ecosystem, and whether the home trial generates regulatory questions in China about liability, data privacy, and safety certification for in-home robotic systems. The technology question -- can these robots do useful things in real homes -- will start to get answered this summer. The commercial question -- who pays, how much, and through what channel -- will take longer.
Frequently Asked Questions
What is X Square Robot and where is it based?
X Square Robot is a Shenzhen-based Chinese startup focused on general-purpose humanoid robots for household use. The company was founded by engineers from leading Chinese AI research institutions and has been building toward home deployment from the outset, rather than starting with industrial applications. As of April 2026, it has raised approximately $580 million across its Series A, A++, and the newly announced Series B.
What is the WALL-B foundation model?
WALL-B is X Square Robot's primary behavioral AI model for household task execution. It handles action planning and physical task completion -- taking a high-level instruction like "prepare breakfast" and breaking it into manipulation steps. It works alongside WALL-OSS, the company's perception model, which handles scene mapping, object recognition, and environment state tracking. Both models run onboard the robot without requiring a cloud connection for core tasks.
Why is Xiaomi investing in a competitor to its own CyberOne robot?
Xiaomi's CyberOne is primarily a demonstration platform and technology showcase -- it's not on a near-term commercial product roadmap for homes. Investing in X Square Robot is a financial hedge and a strategic option. If household humanoids become a large consumer category, Xiaomi benefits from equity in the leading product even if CyberOne doesn't reach market. It also creates ecosystem leverage: X Square robots sold in China could integrate with Xiaomi's smart home platform, which already has tens of millions of connected devices.
How does X Square's home trial plan compare to Western competitors?
X Square is targeting home trials in May 2026 -- one of the most aggressive near-term timelines of any humanoid company globally. Sunday Robotics, the closest US equivalent, is targeting a Thanksgiving 2026 consumer launch for its wheeled Memo robot. 1X Technologies (NEO) has committed to a home beta program in 2026 but has not given a specific month. UniX AI began delivering its Panther humanoid to homes in Suzhou, China in April 2026, making it the current leader in actual home deployments. X Square would be the second Chinese company to reach real homes with a bipedal platform at scale.
What tasks will the home trial robots handle?
X Square has described its household application targets as cleaning, cooking assistance, and elder care support. The May 2026 trials through 58.com will likely focus on a narrower set of high-value, repeatable tasks -- surface cleaning, object retrieval, simple kitchen prep -- rather than attempting the full scope at launch. The trial data will inform which capabilities are reliable enough for broader commercial deployment and which need more training time.
The 12-Month Outlook
X Square Robot's $276 million Series B is the largest single funding event in Chinese household humanoid robotics on record. The company has the capital, the investor network, and a hard public commitment to home trials by the end of May 2026. Whether those trials produce evidence of commercial viability or expose the limits of current household AI will become visible faster than most industry observers expected.
The strategic calculus for Western companies is straightforward: X Square is moving at a pace that assumes the home robotics market is closer than the consensus view. If the May trial produces useful data and a working product, the timeline for household humanoids reaching mass adoption compresses significantly. The relevant comparison is not "when will robots be good enough for homes?" but "which company will have enough real-world home data to train the version that is?"
The Bottom Line: X Square Robot's $276 million raise, backed by Xiaomi, Sequoia China, ByteDance, Meituan, and Alibaba, is the clearest signal yet that China's household humanoid race is accelerating past projections -- and the May 2026 home trial will be the first real test of whether those ambitions are grounded in working technology.
Home trials begin in six weeks. The results will matter beyond X Square -- they'll set the empirical baseline for what household humanoid AI can actually do in 2026, and that data will influence product roadmaps and investor calculus across the entire sector.