# Reachy Mini: An Open Journey Built Together with Hugging Face, Pollen Robotics, & Seeed Studio Source: https://www.youtube.com/watch?v=Lus1HBWY2E8 Recap page: https://rapidrecap.app/video/Lus1HBWY2E8 Generated: 2026-01-06T03:33:49.544+00:00 --- ## Quick Overview The collaboration between Hugging Face, Pollen Robotics, and Seeed Studio resulted in the Reachy Mini, an open-source physical agent designed to bring AI closer to everyone, enabling users to easily customize its behavior using LLMs like GPT and integrating new sensors via its open hardware platform. **Key Points:** - The Reachy Mini is the result of a collaboration between Hugging Face (AI models), Pollen Robotics (design), and Seeed Studio (mass production). - The robot was launched in July, starting as a 3D-printed prototype, and mass production began in early November, selling over 3,000 units quickly. - The open-source nature allows users to fully 3D-print replacement parts and add new sensors/gadgets (like extra arms or MMWave sensors) via the USB port on the host. - The primary interaction method is natural language (voice), supplemented by gesture and motion, making it an accessible interface for non-tech users, seniors, and children. - The robot's API is simple, built with a Python SDK, allowing developers to easily download and push applications from platforms like Hugging Face Spaces. - The goal is for Reachy Mini to become the next generation Human-Computer Interface (HCI), acting as a personal translator and an accessible gateway to AI interaction. - The team is excited about future evolutions, including connecting different LLMs (like GPT) and enabling advanced applications like playing roles or acting as an avatar. ![Screenshot at 00:04: A close-up shot of the industrial pick-and-place machine assembling the Printed Circuit Boards \(PCBs\) for the Reachy Mini, illustrating the transition from prototype to mass production capability achieved through the partnership with Seeed Studio.](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-00-04.jpg) **Context:** This video features an interview between Eric Pan, Founder of Seeed Studio, and Matthieu Lapeyre, Co-Founder & CEO of Pollen Robotics, discussing the journey of developing and launching the Reachy Mini robot. The Reachy Mini is presented as an open-source social robot designed to bridge the gap between complex AI models (like those from Hugging Face) and everyday users, emphasizing its accessibility, customizability, and role in advancing human-computer interaction. ## Detailed Analysis The video documents the successful journey of the Reachy Mini robot, a collaborative effort between Pollen Robotics, Hugging Face, and Seeed Studio, moving from an initial prototype to mass production. Matthieu Lapeyre notes that the project took a long journey, culminating in launching the robot in July and achieving over 3,000 sales quickly by early November, necessitating a rapid transition to mass production in Shenzhen, China. The robot's design, inspired by Ghibli aesthetics, features a simple white shape intended to be customizable, allowing users to write on it or add stickers. A key feature is its open-source nature, supported by a simple Python SDK, which makes it highly accessible for developers, including software engineers and AI developers, to build and share applications via platforms like Hugging Face Spaces. Interaction relies heavily on natural language (voice), gesture, and motion, positioning the robot as the next generation HCI—a personal translator or an accessible gateway to AI, even for non-tech users. The flexibility extends to hardware, with users able to add sensors or actuators via the USB port on the host device. The speakers express excitement for future developments, such as running local LLMs and expanding functionality beyond companion roles into areas like acting or performing plays. ### Development & Production Journey - Launched in July as a 3D-printed prototype - Mass production started early November in Shenzhen, China - Over 3,000 units sold quickly - Rapid transition from prototype to mass production was critical due to a short timeline. ### Open Source & Customization - Built with open source principles - Users can fully 3D-print parts for replacement or extension - Open USB port on the host allows adding new sensors (like cameras or MMWave) and actuators. ### AI Integration & Interaction - Primary interaction via natural language (voice) supplemented by gesture/motion - Integrates with large language models (LLMs) like GPT via the cloud, or potentially local LLMs in the future - Simple Python SDK makes it accessible for developers to create and share apps on platforms like Hugging Face. ### Vision for Reachy Mini - Aims to be the next generation Human-Computer Interface (HCI) - Serves as an accessible gateway to AI for everyone, including seniors and children, who may not use computers directly - Potential for advanced applications like acting/plays, personal translation, and serving as a physical avatar. ### Partnership & Community - Collaboration involved Pollen Robotics (design), Hugging Face (AI models), and Seeed Studio (hardware/production) - The open ecosystem encourages a large community of developers to contribute new applications and features. ![Screenshot at 00:01: Introduction scene showing the two hosts \(Eric Pan and Matthieu Lapeyre\) seated at a table with three Reachy Mini robots.](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-00-01.jpg) ![Screenshot at 00:04: Close-up of the SMT machine assembling PCBs for the Reachy Mini, highlighting the manufacturing process in Shenzhen.](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-00-04.jpg) ![Screenshot at 00:38: Eric Pan and Matthieu Lapeyre identified with their titles: Eric Pan \(Founder of Seeed Studio\) and Matthieu Lapeyre \(Co-Founder & CEO of Pollen Robotics\).](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-00-38.jpg) ![Screenshot at 01:45: A short demonstration clip showing a Reachy Mini with a 3D-printed pumpkin costume interacting with children in a home setting, emphasizing customizability.](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-01-45.jpg) ![Screenshot at 07:54: A collage of social media posts showing community excitement, including one user testing code and another presenting the robot on French TV.](https://ss.rapidrecap.app/screens/Lus1HBWY2E8/00-07-54.jpg)