These days there’s an app for almost every imaginable user and use case, but one thing they all have in common is that they’re centered around one device: the smartphone.
That changes today as Hugging Face, the 10-year-old New York City startup best known as the online space to host and use cutting-edge, open-source AI models, agents, and applications, has launched a new app store for the Reachy Mini, its low-cost ($299) open-source physical robot that debuted in July 2025 (the fruit of Hugging Face’s acquisition of another startup, Pollen Robotics).
The new hugging face Reachi Mini app store already hosts a library of over 200 community-built apps, and Reachi Mini owners will be able to download any of these for free right from the start (unlike smartphone apps, there are no monetization options for app creators on this store yet).
The Reachi Mini App Store will also offer Reachi Mini owners – nearly 10,000 units have been sold since last year – an easy means of creating their own custom apps for the small, stationary desktop robot with built-in camera eyes, speakers and microphone through Hugging Face’s existing, AI-powered agent. "ML Intern."
The importance is not just in the hardware, but also in removing it "robotic" barrier; For the first time, individuals without a background in engineering or coding are shipping functional robotics software in less than an hour.
"Anyone can create apps," Hugging Face CEO and co-founder Clément Delangué said in a video interview with VentureBeat. "My intuition is that more and more [AI] Model makers will release releases on the Reachy Mini as a way to test the robotics capability of new models."
Make robots accessible to common people like PCs and smartphones
A technical hurdle in robotics has historically been the lack of high-quality training data.
While large language models (LLMs) have mastered general-purpose coding by training on huge repositories like Microsoft’s GitHub, the amount of code specific to robotics still remains "Small" By comparison (although Github has by far the largest existing, publicly accessible library of robotics code, with over 17,000 different repositories or "repos" dedicated to the area).
This lack of data means that, until now, AI agents were relatively poor at understanding the physical essence of hardware and firmware requirements.
The solution to hugging face is one agentic toolkit Who acts as a mediator. Rather than forcing the user to learn a specific robotics SDK or master the nuances of the robot’s firmware, the toolkit allows the user to describe the desired behavior in plain English – for example, "Wave when someone says good morning".
An AI agent then handles the heavy lifting: It writes the code, tests it against the robot’s specific constraints, and sends the final package.
"Historically, it has been extremely difficult," DeLongue tells VentureBeat about building robotics applications. "But we’ve worked really hard on this topic, with a mix of open sourcing everything we do, working on the right abstraction for robotics, and making it easy for agents to understand and use."
The platform is model-agnostic, supporting a wide range of major intelligence engines. Users can build apps using Hugging Face’s own ML Intern Agent or leverage external models including GPT-5.5, Cloud Opus 4.6, Kimmy 2.6, Mini Max GM5, and Deep Sig V4 Pro.
For real-time conversations, the official conversation apps use OpenAI Realtime and Gemini Live. By providing these high-level abstractions, Hugging Face subverts traditional "integration week" Robotics works in a process that takes minutes.
Low-cost Richie Mini is a hit
To take advantage of the new Hugging Face Richie Mini App Store, users are encouraged to purchase Richie MiniHugging Face, an adorable desktop robot, launched in July 2025 as an affordable, open-source alternative to existing, commercially available robots like Boston Dynamics, whose infamous Spot robot dog sells for around $70,000. Even Chinese competitors start at $1,900+.
In contrast, the Reachi Mini is priced accessible to hobbyists and developers. It comes in two variants:
- Richie Mini Lite ($299 plus shipping): A tethered version that connects via USB and uses an external computer for processing.
-
Richie Mini Wireless ($449 plus shipping): A standalone version with an on-board Raspberry Pi CM4 and Wi-Fi connectivity.
Of the 10,000 Reachy Mini units sold so far, 3,000 were sold in the past two weeks, Delangue said. Hugging Face expects to ship another 1,000 units within the next 30 days.
Even those who do not have a Reachi Mini can develop apps for it using the Reachi Mini App Store and Reachi Apps, which include a 3D simulation of the robot and its reactions.
app Store Itself is hosted on Hugging Face Hub. It functions like a standard software repository but for hardware behavior:
- Search and Install: Users can find apps, click a button and install them directly to their robot.
-
forkability:every app is "forkable," This means that the user can copy an existing app and ask the AI agent to modify it (for example, "Answer this in French").
-
simulation modeImportantly, the store includes a browser-based simulator. This allows users who do not have a physical Richie Mini to create, test, and play with catalogs in a virtual environment.
Both are moving parts of Hugging Face "le robot" Endeavor – a project that began in 2024 with Hugging Face researchers specializing in robotics and AI who were developing and publishing their own open-source code, tutorials, and hardware on the web to make robotics development more accessible to a broader audience.
And unlike GitHub, which is designed for a developer audience, Hugging Face Reachy Mini App Store is designed for robot owners and users who have no technical experience or training.
Continuing the open-source ethos and practices
Hugging Face’s strategy is based on the belief that closed-source hardware and software are "almost impossible" For large scale construction.
DeLongue says closed systems prevent training of agents and limit the community’s ability to innovate. As a result, the entire Reachi Mini platform is open-source.
This open licensing model has two primary implications for the ecosystem:
- rapid development: Because the code is public and integrated with the Hugging Face ecosystem "blank space," Hugging Face feature for a host of AI-powered web apps Launched in 2021, agents can more easily learn how to interact with the hardware.
-
community sovereignty:The apps are not locked behind any proprietary walls. Currently, all 200+ apps on the store are free, although the foundation of the platform is a work in progress "free space" Allows creators to potentially monetize their work in the future.
"At the moment, all apps are free," DeLongue noted. "It’s flexible, it’s constructed [Hugging Face] Spaces, so at some point people are probably going to pay them."
Robotics enters its accessible hobbyist era
Hugging Face’s Reachy Mini App Store is launching with 200 apps already available.
So who created them, and how did they create it without this platform already existing?
DeLongue told VentureBeat even more. 150 different creators have contributed to the store, most of whom had never written a line of robotics code before.
Yet, they have been able to do so thanks to Hugging Face’s ML interns and Github. The new Hugging Face Reachy Mini App Store now houses tools and existing apps in one place for easy access.
DeLongue was particularly keen to highlight one of VentureBeat’s early Reachy robotics app developers: Joel Cohen, a 78-year-old retired marketing executive.
Cohen, who suffers from color-blindness and has no technical background, spent two weeks assembling his Reachy Mini Light (a task that typically takes three hours). Despite these physical challenges, they used an AI agent to create this "Vice President of Future Thinking" Facilitator for your Zoom-based CEO peer groups. The app enables the robot to:
- Greet 29 members by name.
-
Fact-checking discussions in real time.
-
Summarize main topics and emphasize surface level answers.
"I created this by describing what I wanted in simple English," Cohen said in a press release provided to VentureBeat ahead of the launch. "No SDK. No robotics background. no developer experience".
Other community-driven applications include:
- emotional damage chess: A robot that plays chess and makes fun of the user’s mistakes.
-
call home: An anti-procrastination tool that detects when a user picks up their phone and tells them to get back to work.
-
language teacher: a physical companion who listens to speech and corrects pronunciation.
-
F1 race commentator: A deskmate who calls Formula 1 races live as they happen.
DeLongue himself tells VentureBeat that in just a few hours, he created an app for his Reachy mini robot in the Hugging Face Miami office, so the robot could act as a receptionist.
“It basically does facial recognition to figure out when you get to the office, and then it sees you and engages you," Delangu related. "It is written, ‘Hey, welcome to the office. ‘Who have you come here to see?’ Then it sends me a message: ‘Carl has just arrived at the office. He has come here to meet you, and for these reasons.’ This little one serves as my welcome booth at the office, and it took me less than two hours to build.
Even for an experienced founder and developer like Delangeau, building apps for robots was out of the question until the combination of Reachy Mini and ML Intern.
“For me, it would have been impossible," said the hugging face CEO. "If you weren’t a robotics developer, it might have been impossible, or it might have taken months."
Democratization of robotics
The launch of the Agentic App Store signals a fundamental change in the way we interact with machines. For sixty years, this field was beset by the need for deep technical expertise.
By combining low-cost open hardware with the reasoning capabilities of modern AI agents, Hugging Face is moving toward a future where hardware is a commodity and behavior is limited only to what the user can describe.
As Delangu noted during the launch, the goal was to provide a platform for those "Want to get into robotics but don’t have the hardware or skills".
Now with about 10,000 robots "in the wild" And a growing store of agent-written apps, the Reachy Mini has become the most widely deployed open-source desktop robot in history.
The question now is not how to build robots, but what we will ask them to do now that the gate is open.
<a href