Feather Robotics Builds the Android of Robotics
Feather Robotics, founded in 2024, is creating a modular humanoid robot platform that developers can customize for tasks like cooking and lab cleanup. With a $30,000 price tag and support for multiple AI models, the startup is positioned as a U.S. alternative to Chinese competitors and aims to buil…
By Felo News Desk · Published
Feather Robotics, a startup founded last year, is taking a different tack on the robotics race. While companies like Tesla and Figure are still working on a complete, general‑purpose robot, Feather is delivering a modular humanoid platform that developers can buy and program today. The result is a robot that can be re‑configured for a range of tasks—from cooking in a Japanese restaurant to cleaning up a science laboratory—without waiting for a future “ChatGPT moment” in robotics.
What Feather Is Building
Feather’s hardware is designed around a modular architecture. The base chassis can be fitted with arms of varying lengths, different sensor suites, and interchangeable end‑effectors. This flexibility means that a single robot can be adapted for a wide spectrum of use cases. The company has already begun selling units, reporting over $1 million in revenue, and it claims that its robots are currently deployed in restaurants in Japan and in research labs.
On the software side, Feather’s platform is agnostic to the AI model provider. Whether a developer chooses Generalist, Skild, Physical Intelligence, or any other leading robotics AI, the robot can run the model on its onboard hardware. This openness is a key selling point for companies that want to experiment with different AI approaches without being locked into a proprietary ecosystem.
Why Feather Is Different
Feather’s co‑founders, Hoa Mai and Parsa Bakhtiari, bring a mix of experience from the automotive and robotics worlds. Mai, who sold his previous humanoid startup to 1X in 2025, teamed up with Bakhtiari—a former Tesla Model 3 engineer who once reported directly to Elon Musk—to create a product that could be shipped now. They were motivated by the fact that most successful hardware companies, like Nvidia and Apple, started with a working product and added complexity over time. “You can’t buy a Tesla robot today and develop on top of it,” Mai told TechCrunch. “We realized that this is not how most companies became successful.”
Feather’s pricing strategy also sets it apart. At $30,000 per unit, the robot is roughly half the price of Unitree’s H2 Edu, a popular Chinese model that has been restricted from entering the U.S. market. According to Gradient Ventures general partner Darian Shirazi, the lower price point opens a large market. “You would hire a laborer for $50,000 to $60,000 a year, you’d have to train them, manage HR, and so on. You can now buy a Feather robot to do that,” Shirazi said.
The Ecosystem Vision
Feather is not just selling a robot; it’s building an ecosystem. The company believes that the real value in robotics will come from the thousands of application companies that will develop around its platform. Mai says, “If we think about the market size today for physical AI companies, it’s very small, but if you think about how many physical AI application companies will probably exist in five years, we expect it to be in the thousands.”
To support this vision, Feather has kept its operations lean. The startup has spent only a fraction of its $7.6 million pre‑seed round, allowing it to focus on product development and customer success. The company’s early customers—though not named publicly—include restaurants in Japan and science labs, indicating that the platform is already proving useful in real‑world settings.
What’s Next for Feather
After a year of field testing, Feather says it has resolved most of the issues that arose during early deployments. The company is preparing for a larger product launch, aiming to scale production and expand its customer base. With a clear modular design, a competitive price, and an open‑AI ecosystem, Feather is positioning itself as a U.S. alternative to Chinese robotics firms and as a platform that could power a new generation of physical AI applications.
As the robotics market continues to evolve, Feather’s approach of delivering a deployable, customizable robot now may prove to be the catalyst that accelerates the industry toward the long‑awaited general‑purpose robot.
Key facts
- Feather offers a modular humanoid robot that can be customized for tasks like cooking and lab cleanup.
- The robot runs AI models from any major robotics AI provider, ensuring flexibility.
- At $30,000, Feather’s robot is half the price of comparable Chinese models and is positioned as a U.S. alternative.
- The company has already sold over $1 million in revenue and is preparing for a larger product launch.
- Feather’s founders believe the real value lies in building an ecosystem that could support thousands of application companies in the next five years.
Why it matters
Feather’s modular, affordable platform could democratize robotics, allowing companies to deploy robots for specific tasks today instead of waiting for a future breakthrough. This approach may accelerate the adoption of physical AI across industries.
Frequently asked questions
What makes Feather’s robot modular?
Feather’s base chassis can be fitted with different arm lengths, sensor suites, and end‑effectors, allowing the same robot to perform a wide range of tasks.
Can Feather’s robot run any AI model?
Yes, the platform is agnostic to the AI model provider, supporting models from Generalist, Skild, Physical Intelligence, and others.
How does Feather’s price compare to competitors?
At $30,000, Feather’s robot is roughly half the price of Unitree’s H2 Edu, making it more affordable for many businesses.
What industries are already using Feather’s robots?
Early deployments include restaurants in Japan and science laboratories, though the company does not disclose all customer names.
Sources
- [1] techcrunch.com — originally reported as “Meet Feather, the startup building the 'Android of robotics' for developers”
