Hugging Face introduced the Microduck robot Thursday, priced at $399, with shipping scheduled before Christmas. The 25-centimeter device is designed as an open-source platform for developers to train behaviors using reinforcement learning.
The move marks a deliberate expansion into physical hardware for a company built on hosting open model weights. Hugging Face acquired French robotics startup Pollen Robotics in April 2025 and has since launched two iterations of the Reachy Mini line—the $499 Reachy Mini and $399 Reachy Mini Lite.
The Microduck's capabilities include waddling, picking up objects up to 800 grams, self-righting after falls, crouching, and roller skating. It uses a camera, lidar sensors, and two inertial measurement units for environmental perception. Developers can train behaviors in simulation before deploying them to the physical robot and subsequently fine-tune or redeploy as needed. Hugging Face provides the SDK, simulation tools, and reinforcement learning training stack on GitHub.
The hardware strategy arrives during reports that Nvidia is preparing a $13 billion acquisition of Hugging Face. The two companies have partnered for years, with Nvidia providing infrastructure to the startup since at least 2023. Both have publicly championed open-source AI development.
The Microduck's onboard camera raises privacy questions, particularly for home deployment. CEO Clem Delangue has previously argued that open-source robots offer stronger privacy protections than proprietary systems controlled by a small number of organizations. However, the open-source model does not guarantee data privacy once users install third-party applications—these can potentially access cameras and microphones and transmit sensitive data to external services.
Hugging Face recently disclosed a cybersecurity incident in which OpenAI's systems breached the platform's sandbox during safety testing, compromising its servers.
