In an announcement that turned more than a few heads, Unitree Robotics opened UniStore — the world's first humanoid robot app store — to the public last week. Think the App Store, but for downloading boxing moves, dance routines, and Kamehameha blasts onto a walking, talking robot.

The Hangzhou-based startup quietly flipped the switch on May 7, making 24 motion apps available for its G1 humanoid robot with one-tap installation from a phone app. Bruce Lee's Jeet Kune Do, Michael Jackson's dance moves, Mantis Boxing, and even a Dragon Ball Z-inspired power-up sequence are all available for download.

It sounds like a joke. It is not.

🔥 WHAT HAPPENED

Unitree opened UniStore (unistore.unitree.com) on May 7, 2026 — a shared application platform for humanoid robot tasks and motions. It is, by the company's own account, the first of its kind anywhere in the world.

Here's what matters:

The platform is currently compatible with the G1 humanoid robot, with minimum system version requirements listed on the site. Users browse an action library of 24 distinct executable motions, tap install from their phone, and the robot executes them autonomously.

The current lineup includes boxing drills, Charleston dancing, Jackson-inspired moves, Kamehameha postures, and basic locomotion patterns like striding and obstacle stepping. The company also released a public SDK for developers to build and submit their own motion packages.

The timing is deliberate. UniStore launched just one week after Unitree unveiled the R1 series dual-arm humanoid robot — a $3,949 entry-level platform designed for upper-body manipulation with 15 to 31 degrees of freedom. The R1 supports OTA updates and optional Nvidia compute modules for higher-level AI workloads.

Unitree also opened its first direct-sale retail store in Beijing's Wangfujing commercial district the same week, signaling a serious push into consumer markets.

🧠 WHY THIS MATTERS

This is not a toy store. UniStore represents a fundamental shift in how robotics companies think about their products.

The traditional robotics model is closed. You buy a robot, you get whatever software and skills it ships with. If you want it to do something new, you either hire a team of engineers or buy a new robot. Unitree is flipping that model on its head by treating robot skills as downloadable apps — a smartphone business model for a hardware industry that has never had one.

Think about what the iPhone App Store did for smartphones: it turned a hardware device into a platform. Before 2008, phones did what manufacturers decided they should do. After the App Store, phones could do anything a developer could imagine. Unitree is betting that the same dynamic will unlock the humanoid robot market.

The numbers back this up. Unitree shipped more humanoid robots than any other company in 2025, generating 1.708 billion yuan ($235 million) in revenue — a 335% year-over-year surge. Humanoid robots became the company's largest revenue source. They are preparing for a 4.2 billion yuan IPO on Shanghai's STAR Market.

This is not a startup playing around. This is a company with market leadership trying to build the platform that the entire industry runs on.

📊 DEEP DIVE

The Digitimes analysis that ran today makes a compelling case that Unitree is playing a different game than competitors like Tesla, Boston Dynamics, or Figure. Those companies are in a hardware arms race — who can build the most capable robot body. Unitree seems to be focusing on something else: who can build the most compelling robot ecosystem.

The parallel to smartphones is instructive. In 2007, Nokia had the best hardware. BlackBerry had the best keyboard. Apple had neither — but it had an app store. Ten years later, the app store won everything.

UniStore works through a mobile app interface. Users browse categories, tap to install, and the robot receives the motion package via OTA update. There is no coding required. No robotics expertise needed. If you can install Candy Crush, you can teach your robot to dance.

The SDK opens the other side of the marketplace. Third-party developers can build motion packages, submit them for review, and distribute them through the store. This is the flywheel that made the App Store unstoppable: more developers attract more users, which attract more developers.

Unitree's broader hardware lineup shows how the strategy connects: the G1 (mid-range research robot at $13,500-$27,000), the H1/H2 (full-size industrial robots starting at $29,900), the B2 (quadruped), and the Go2 (consumer robot dog). Every single one of them can now run apps from UniStore.

The R1, priced aggressively at $3,949, is designed to be the iPhone SE of the humanoid robot world — cheap enough that developers and hobbyists can afford one, capable enough to run real applications.

⚠️ THE CATCH

A store with 24 apps that mostly consists of dance moves and kung fu poses is not a revolution yet — it is a proof of concept. The real question is whether the app ecosystem can cross the chasm from novelty to utility.

The G1, at $13,500-$27,000, is still expensive for most consumers. The R1 at $3,949 is more accessible, but it focuses on upper-body manipulation — not the full locomotion capabilities that make humanoid robots compelling. The audience for a $4,000 torso robot that can perform Meme Boxing is limited.

Developer traction is unproven. The App Store worked because Apple had millions of users willing to pay for apps. Unitree has thousands of robot owners. The economics of building a business on a motion app store for a niche hardware platform are unclear.

There is also a safety question that nobody is talking about yet. If a robot can download a new motion package and execute it autonomously, what happens when a bad app is uploaded? Unitree has review processes, but the App Store has proven that review processes are imperfect. A robot moving with unknown intent in a home or workplace is a fundamentally different risk profile than a smartphone displaying a buggy app.

And then there is the competitive landscape. Agibot, Tesla, and UBTECH are all racing on hardware. If any of them builds a more capable robot body at a lower price point, the platform strategy only works if developers and users are locked in. That lock-in is not guaranteed when the hardware market is still so fragmented.

🎯 WHAT HAPPENS NEXT

The next 90 days will tell us whether UniStore is the beginning of something huge or a clever gimmick.

Key milestones to watch:

First, third-party app submissions. If developers start building and publishing utility apps — warehouse navigation, inspection routines, household assistance — it signals real platform momentum. If the store remains a collection of dance moves, it is a novelty.

Second, the R1 adoption rate. The $3,949 price point is designed to put humanoid robots in the hands of developers worldwide. If R1 sales take off, the addressable market for UniStore explodes. If they don't, the store remains a niche within a niche.

Third, the IPO. Unitree is targeting a 4.2 billion yuan listing on Shanghai's STAR Market. The UniStore launch adds a software narrative to what would otherwise be a pure hardware pitch. Investors love platform stories. How they value Unitree will signal the market's conviction.

Fourth, international expansion. Unitree is already selling through cross-border e-commerce platforms like AliExpress. A UniStore that works in multiple languages and markets transforms the company from a Chinese hardware maker into a global robotics platform. That is the difference between a $1 billion company and a $100 billion one.

🧩 BIGGER PICTURE

The humanoid robot industry is at a fork in the road. One path leads to a world where robots are specialized machines bought by corporations for specific tasks — the industrial robot model that has dominated for 50 years. The other path leads to a world where robots are general-purpose platforms that can learn new skills through software — the smartphone model.

Unitree is betting everything on the second path. UniStore is the infrastructure for that bet. The app store model worked for phones because it solved a fundamental problem: hardware is finite, but software is infinite. The same logic applies to robots. A robot body is a fixed piece of machinery. But what it can do — what skills it has — is only limited by the software ecosystem around it.

The irony is that this is exactly the playbook China used to dominate the smartphone industry. Chinese manufacturers started with cheap hardware, built volume, then layered on software ecosystems that competed on utility rather than spec sheets. Unitree appears to be running the same play for robots.

Whether it works depends on whether the world actually wants humanoid robots doing kung fu in their living rooms. But that was also the question about smartphones in 2008. The answer turned out to be: not for the apps they had — for the apps they would eventually have.

UniStore is not about Mantis Boxing. It is about proving that a robot can learn new tricks without being rebuilt. That is the idea that could change the industry.

Sources: CnTechPost, Pandaily, Digitimes, KraneShares, Forbes, RobotToday