Feather Robotics launches $29,990 modular wheeled humanoid platform for third party developers

Feather Robotics Inc. announced on September 24, 2026, that it raised a $7.6 million pre-seed round led by Gradient and launched a modular wheeled bimanual humanoid robot priced at $29,990. The East Palo Alto, California, company, founded in 2025 by Hoa Mai and Parsa Bakhtiari, says the system is for third-party developers who want to build physical AI applications without a vertically integrated hardware stack. Feather has already generated more than $1 million in revenue.

The startup aims to build the Android of robotics. TechCrunch reported on September 24, 2026, that co-founder Hoa Mai said a developer cannot buy a Tesla robot today and develop on top of it. Mai previously founded Kind Humanoid, which was acquired by 1X. Bakhtiari is a former Tesla Model 3 engineer who reported directly to Elon Musk and helped scale Model 3 production during Tesla's 2018 production crisis.

Feather's robot does not walk on legs. It uses a wheeled, holonomic base that moves in multiple directions. The company says this improves stability and energy efficiency while reducing mechanical complexity. The robot is bimanual and modular from the ground up, so developers can adjust components such as arm lengths. Each system ships with a software development kit, and the hardware can run models from Nvidia, Skild, and Physical Intelligence.

Pricing is central. Feather's system is available to order online for $29,990, well below listed prices for comparable systems: about $68,000 for Unitree's H2 Edu, about $135,000 for Agibot's G2, and about $83,000 for Dexmate's Vega. Feather says its robot costs one-half to one-third of those Chinese counterparts. It claims a 10-hour battery life, roughly three times that of Unitree's H2, and about 1 meter of reach, roughly 50 percent longer than the average human arm.

Key Facts

The $7.6 million pre-seed round was led by Gradient Ventures, with participation from Builder Capital, Geometry, SEED Innovations, and Virgo VC. Investegate reported on September 25, 2026, that SEED Innovations Limited confirmed the round was a Simple Agreement for Future Equity, or SAFE. SEED contributed $1.0 million, and that investment will convert into equity at the next qualifying funding round or liquidity event, subject to a valuation cap of $60 million.

The AI Insider reported on September 26, 2026, that Feather's wheeled bimanual robot is priced at $29,990, offers up to 10 hours of battery life, has about 1 meter of reach, and includes an SDK. The company built its first prototype in two months, fulfilled its first customer order within nine months, and has generated more than $1 million in revenue. By comparison, Feather says Figure took 31 months to fulfill its first order.

Customers include companies in manufacturing and food service, including a manufacturer with $4 billion in annual revenue. Feather is not disclosing its customers by name, but it said its robots are working as cooks in restaurants and cleaning up science labs. Forbes reported on September 25, 2026, that Feather officially launched a vaguely humanoid robot on wheels at $29,990, backed by the $7.6 million pre-seed, and that it already has more than $1 million in revenue according to the founders.

TechCrunch reported on September 24, 2026, that Gradient general partner Darian Shirazi said Feather does not have direct competitors in the United States. Shirazi breaks the landscape into domestic household robots such as Sunday, general-purpose heavyweights such as Figure and Tesla, and Feather as the only U.S. startup developing a modular humanoid platform. He argued the market at that price could be very large, saying that a company might hire a laborer for $50,000 to $60,000 a year, and that it can now buy a Feather robot to do that work.

Feather has spent only a fraction of its pre-seed funding, according to TechCrunch. The full launch announcement was dated September 24, 2026, at 10:00 a.m. Pacific Time. The robot can be ordered online. Feather says its bimanual design uses a wheeled, holonomic base, with a payload designed to carry loads comparable to an adult human.

Analysis

The bigger picture here is that Feather is not trying to win the humanoid race by building the most capable bipedal machine. It is trying to become the default development layer for physical AI. By pricing a modular, wheeled, bimanual robot at $29,990 and shipping it with an SDK, Feather is lowering the cost of experimentation for third-party developers and small companies that cannot afford or access vertically integrated systems from Tesla or Figure. That is closer to selling picks and shovels than to selling a finished robot workforce.

The strategy has clear logic. The AI Insider reported on September 26, 2026, that Feather frames its hardware as an alternative both to vertically integrated humanoid systems that are difficult for outside developers to modify and to lower-cost research platforms that may not be designed for commercial deployment. Feather sits deliberately between those two poles. Its wheeled base avoids some of the hardest mechanical problems in bipedal locomotion, which reduces complexity and likely improves reliability. The trade-off is that wheels limit terrain, but for many indoor commercial tasks in restaurants, labs, and factories, a wheeled base may be sufficient. Feather's claim of 10 hours of battery life and 1 meter of reach suggests it is targeting practical, multi-hour work rather than short demos.

What this really means is that the value in humanoid robotics may increasingly shift from hardware to software and ecosystem. If developers can buy a capable modular robot for under $30,000 and run models from Nvidia, Skild, or Physical Intelligence, then the winning platform may be the one that attracts the most applications, not the one with the most human-like body. Gradient's Darian Shirazi told TechCrunch on September 24, 2026, that the winners in physical AI will provide a seamless hardware-and-software platform. Feather's bet is that an open, modular approach can create a network effect similar to Android in smartphones.

There are risks. Feather faces competition from Chinese manufacturers such as Unitree, Agibot, and Dexmate, even if Shirazi says there is no direct U.S. competitor for a modular humanoid platform. Those companies have listed prices that are higher than Feather's, but they also have established hardware supply chains. Feather's advantage is its developer focus and its early revenue, which is rare for a pre-seed hardware startup. More than $1 million in revenue before a large product launch gives the company real field data. Feather says it has resolved almost all issues found during a year of field testing, and that operational experience may matter more than raw specifications.

Why It Matters

Feather's launch matters because it tests whether small teams can rewrite the economics of humanoids. Forbes reported on September 25, 2026, that the launch is part of a trend in which small teams are rewriting the economics of humanoids, offering a deployable, developer-oriented platform at a price point far below competitors. If Feather succeeds, it could prove that a startup with a $7.6 million pre-seed and a two-month prototype cycle can compete in a field dominated by companies with far larger balance sheets.

It also matters for labor economics. Shirazi's comparison to hiring a laborer for $50,000 to $60,000 a year frames the robot as a capital purchase that could substitute for some repetitive or dangerous work. At $29,990, the payback period could be attractive for tasks that run multiple shifts. Feather's customers already include a $4 billion revenue manufacturer, restaurants, and science labs. These are commercial environments where reliability and support logistics matter. The company says its robots are working as cooks and cleaning up science labs, which suggests early deployments are in structured indoor spaces.

The broader significance is strategic. The United States has few domestic options for developers who want to build on humanoid hardware without partnering with a vertically integrated giant. Feather argues that this gap has forced independent developers to choose between expensive, closed systems and cheap platforms that may lack commercial reliability. By offering a modular robot with an SDK and a $29,990 price, Feather is trying to create a middle path. If that path works, it could become the foundation for a new layer of physical AI applications, from restaurant automation to laboratory support.

Next Up

Feather says it is preparing for a big product launch after resolving almost all issues found during a year of field testing. CEO Hoa Mai told TechCrunch on September 24, 2026, that the company has been selling small quantities of the robots and is now getting ready for a larger launch. The robot is already orderable online at $29,990. SEED Innovations' $1.0 million investment will convert into equity at the next qualifying funding round or liquidity event, subject to a $60 million valuation cap.

The company has spent only a fraction of its $7.6 million pre-seed, according to TechCrunch. That leaves room to expand manufacturing, support, and developer outreach as it moves from early field testing to broader commercial deployment. The key question is whether Feather can turn its early revenue and modular developer pitch into a durable platform business. If it can, the Android of robotics may move from a tagline to a real market position.