From Science Fiction to Factory Floors: The Rise of Humanoid Robots
Humanoid robots — machines designed to look and move like humans — have been a staple of science fiction for decades. But over the last few years, they've stepped off the screen and into real warehouses, hospitals, living rooms, and assembly lines. What was once a niche research field has turned into one of the most competitive industries in the world, with dozens of companies racing to build the first truly mass-produced general-purpose humanoid robot.
In this article, we break down the top 10 humanoid robot manufacturers in the world as of 2026 — covering their history, what makes them unique, famous robot models, key breakthroughs, commercial operations, competitive edge, future plans, and expected production volumes.
1. Tesla (Optimus / Tesla Bot)
Country: USA | Founded: 2021 (project announced) | Founder: Elon Musk
History: Tesla announced the Optimus (also called Tesla Bot) at its AI Day event in August 2021. By September 2022, a crude prototype was walking on stage. Since then, development has accelerated rapidly, leveraging Tesla's expertise in electric motors, batteries, computer vision, and AI chips — all from its electric vehicle business.
Famous Models: Optimus Gen 1 (2022), Optimus Gen 2 (2023), Optimus Gen 3 (2024/2025)
Uniqueness: Optimus uses the same FSD (Full Self-Driving) computer, cameras, and AI inference hardware found in Tesla vehicles. This gives Tesla a massive supply chain advantage — Optimus components can be manufactured on existing car production lines.
Pros: Unmatched supply chain & manufacturing scale; vertical integration (motors, batteries, silicon); access to real-world AI data from millions of Teslas; Elon Musk's ability to attract top engineering talent.
Cons: Still not in commercial production; Musk's timelines have been consistently over-optimistic; lacks a commercial deployment track record; competition is catching up fast.
Key Breakthroughs: Self-correcting walking algorithms; object manipulation using Tesla vision; end-to-end neural network control (no hand-coded routines).
Commercial Operation: None yet — Optimus is still in development. Tesla has demonstrated it performing simple factory tasks (moving boxes, watering plants) in controlled settings.
Target Industries: Manufacturing, logistics, warehouse operations, and eventually household service.
Expected Production: Musk has claimed "millions" eventually. Initial production likely 1,000–10,000 units in 2026–2027.
2. Figure AI
Country: USA | Founded: 2022 | Founder: Brett Adcock
History: Figure AI was founded in 2022 by serial entrepreneur Brett Adcock (also founded Archer Aviation and Hired [formerly Vettery]). The company raised $70M in May 2023, then secured a massive funding round including Microsoft, OpenAI, NVIDIA, Jeff Bezos, and Intel. As of late 2025, Figure AI was valued at an estimated $39 billion.
Famous Models: Figure 01 (2023), Figure 02 (2024), Figure 03 (2025–2026)
Uniqueness: Figure's partnership with OpenAI gives its robots cutting-edge vision-language models. Their Helix AI model (vision-language-action) can control two robots simultaneously from a single neural network — a breakthrough in multi-robot coordination.
Pros: Deep-pocketed investors ($39B valuation); strong AI partnership with OpenAI; rapid iteration (3 generations in 3 years); demonstrated commercial deployments.
Cons: Extremely high burn rate; limited actual production volume so far; very young company with unproven manufacturing at scale.
Key Breakthroughs: Helix vision-language-action model for dual-robot control; Figure 03 appeared at a White House event in 2026 showcasing household tasks.
Commercial Operation: Deployed at BMW's Spartanburg plant for automotive manufacturing testing — one of the few real commercial pilot programs in the industry.
Target Industries: Automotive manufacturing, logistics, warehousing, retail, and eventually household service.
Expected Production: Plans to scale to 10,000+ units per year by 2027–2028, with a 50,000-unit factory discussed.
3. Boston Dynamics
Country: USA | Founded: 1992 (MIT spin-off) | Owner: Hyundai Motor Group (since 2020)
History: Boston Dynamics is the original pioneer of advanced robotics. Founded as an MIT spin-off, the company spent decades building research robots funded by DARPA contracts. Its hydraulic-powered robots (BigDog, LS3, Atlas) were legendary for agility. Hyundai Motor Group acquired Boston Dynamics in December 2020 for $1.1 billion.
Famous Models: BigDog (2005), LS3 (2012), Atlas (2013–present), Spot (2015–present), Stretch (2021–present)
Uniqueness: Unmatched mobility and dynamic control. Boston Dynamics robots can do parkour, backflips, dance routines, and navigate the most challenging terrain of any robot ever built. Their whole-body manipulation approach is the gold standard.
Pros: Decades of R&D; world-leading locomotion algorithms; Hyundai's manufacturing backing; proven commercial products (1,000+ Spot units deployed).
Cons: Atlas is still a research platform (not commercially available); extreme cost (Spot starts at $74,500); hydraulic Atlas too complex for mass production; new electric Atlas (2024) is unproven commercially.
Key Breakthroughs: Atlas doing backflips (2017); parkour (2018–2023); whole-body manipulation with heavy objects; switch from hydraulic to fully electric Atlas (2024).
Commercial Operation: Spot sells commercially for industrial inspections (oil & gas, construction, power plants) with 1,000+ units sold. Stretch deployed for warehouse unloading. Atlas transitioning from research to commercial with Hyundai.
Target Industries: Industrial inspection (Spot), warehouse logistics (Stretch), automotive manufacturing (Atlas + Hyundai), construction.
Expected Production: Spot produced in hundreds per year. Electric Atlas aims for low thousands within 2–3 years of commercial launch.
4. Agility Robotics
Country: USA | Founded: 2015 (Oregon State spin-off) | CEO: Peggy Johnson
History: Agility Robotics spun out of Oregon State University's Dynamic Robotics Laboratory in 2015. Founders Jonathan Hurst and Damien Shelton developed Cassie — a bipedal research platform. In 2019, they introduced Digit, the first humanoid robot designed specifically for commercial logistics. Amazon invested early and placed the first major pilot order.
Famous Models: Cassie (2017), Digit (2019–present)
Uniqueness: Agility's "legs-first" philosophy — they believe walking and balancing are the hardest problems, so they solved those first before adding an upper body. Digit is the first humanoid robot commercially available for purchase (starting at ~$250,000).
Pros: First company to sell humanoid robots commercially; real customer deployments (Amazon, GXO, ERCO); purpose-built for logistics; working manufacturing facility "RoboFab" in Oregon.
Cons: High price ($250K); limited upper-body strength; not suitable for heavy manufacturing; battery life (~2 hours); still in early commercial phase.
Key Breakthroughs: First commercial humanoid robot sale; autonomous tote-moving in warehouses; integration with Amazon's robotic fleet; outdoor walking capability.
Commercial Operation: Digit deployed in real warehouses — Amazon pilots in Seattle, GXO logistics centers, ERCO facilities. RoboFab manufacturing facility opened in Salem, Oregon in 2024.
Target Industries: Warehouse logistics, material handling, e-commerce fulfillment.
Expected Production: RoboFab capacity is 10,000+ units per year. Agility plans to reach thousands by 2026–2027.
5. Unitree Robotics
Country: China | Founded: 2016 | Founder: Wang Xingxing
History: Unitree started as a quadruped robot company (competing with Boston Dynamics' Spot at a fraction of the cost). Their first humanoid robot, the H1, was unveiled in 2023. By 2024 they released the G1 humanoid for just $16,000 — shocking the entire industry with the price point.
Famous Models: Go1 (quadruped), B2, H1 (2023), G1 (2024), H1-2 (2025)
Uniqueness: Unitree's laser focus on cost reduction and aggressive pricing. The G1 at $16,000 is the cheapest humanoid robot on the market by a massive margin (Digit costs ~$250K, Spot costs $75K). They've proven humanoid robots can be manufactured at consumer-level prices.
Pros: Aggressive pricing ($16K for G1); fast iteration; proven quadruped mass-production experience; strong supply chain in China; constant tech media attention.
Cons: Lower reliability than US competitors; software/AI less advanced than Figure or Tesla; some test footage appears cherry-picked; controversial ties to Chinese government funding.
Key Breakthroughs: $16K humanoid robot (G1); H1 backflip demonstrations; H1-2 with advanced dexterity; impressive running speed.
Commercial Operation: G1 and H1 available for purchase today online through Unitree's store. Primarily sold to research labs, universities, and early adopters.
Target Industries: Education, research, entertainment, light industrial.
Expected Production: Thousands per year — already producing in volume with Chinese manufacturing muscle.
6. 1X Technologies
Country: Norway / USA | Founded: 2014 | CEO: Bernt Øivind Børnich
History: 1X Technologies started in Norway, developing wheeled humanoid robots for security and caregiving. The company exploded onto the scene when OpenAI invested (along with Tiger Global) in a 2023 Series A2 round. They now have a manufacturing facility in Hayward, California, with main operations in Moss, Norway.
Famous Models: REV1 (wheeled, security), EVE (wheeled, enterprise), NEO (bipedal, 2024–2025)
Uniqueness: 1X takes a practical, safety-first approach. EVE uses wheels instead of legs (faster, cheaper, more stable) for enterprise tasks. NEO is their first bipedal robot — designed for homes — using compliant, soft-touch materials for safe human interaction.
Pros: OpenAI backing and AI integration; practical safety-first design (soft materials, wheels); EVE already commercially deployed in security; manufacturing facility in California.
Cons: Wheeled robots can't climb stairs (not "truly" humanoid); NEO not yet in mass production; limited capital compared to Tesla/Figure/Boston Dynamics.
Key Breakthroughs: NEO's compliant actuation system (soft, spring-loaded joints for human safety); EVE's autonomous patrol and monitoring capabilities.
Commercial Operation: EVE deployed in commercial security and facility monitoring roles. 1X has a manufacturing facility in Hayward capable of producing thousands of units.
Target Industries: Security & surveillance (EVE), home assistance (NEO), elderly care, light industrial.
Expected Production: Thousands per year — initial capacity for 10,000 NEO units at Hayward factory.
7. UBTECH Robotics
Country: China | Founded: 2012 | Founder: Zhou Jian
History: UBTECH is one of China's largest consumer robotics companies, best known for its educational robots (Alpha Mini, Yanshee, uKit) used in schools worldwide. In recent years, they've pivoted to humanoid industrial robots with the Walker series.
Famous Models: Alpha Mini (educational), Yanshee, Walker (2018–present), Walker S (2024)
Uniqueness: UBTECH has a dual-market strategy: it already makes money selling educational robots to schools worldwide, while investing those profits into humanoid R&D. The Walker S is deployed in NIO electric vehicle factories — one of the few real humanoid deployments in automotive manufacturing.
Pros: Profitable educational robot business; funding from Chinese government initiatives; real factory deployments (NIO); strong IP portfolio; mass-production experience with consumer robots.
Cons: Walker's capabilities lag behind US/European competitors; mostly focused on China domestic market; less advanced AI/software; limited Western market access.
Key Breakthroughs: Walker's bipedal walking and object manipulation; Walker S deployed in NIO EV factory; integration with Chinese smart manufacturing initiatives.
Commercial Operation: Walker S deployed at NIO's EV factories for assembly line tasks. Strong educational robot business in 50+ countries.
Target Industries: Automotive manufacturing (NIO), education, smart manufacturing, logistics.
Expected Production: Infrastructure for tens of thousands of units (based on educational robot production). Walker S likely hundreds to low thousands in 2026.
8. Sanctuary AI
Country: Canada | Founded: 2018 | CEO: Geordie Rose
History: Founded by Geordie Rose — the same entrepreneur behind D-Wave Systems (quantum computing) and Kindred Systems (robotics). Sanctuary AI is on a mission to build the world's first "human-like intelligence" in a robot body, with deep focus on dexterity and tactile sensing.
Famous Models: Phoenix (2023–present)
Uniqueness: Sanctuary focuses on human-like hands and dexterity above all else. Their Carbon haptic sensing system gives Phoenix tactile feedback — the robot can literally feel what it touches. They believe general-purpose labor requires human-level manipulation, not just walking.
Pros: Best-in-class dexterity and tactile sensing; strong AI focus; Canadian government support and research grants; teleoperation-based data collection accelerates learning.
Cons: Phoenix slower and less mobile than competitors (still refining walking); hasn't demonstrated mass-manufacturing capability; limited funding compared to Tesla/Figure; small team.
Key Breakthroughs: Carbon haptic tactile sensing system; Phoenix's 20-degree-of-freedom hands; autonomous retail task completion (stocking, folding, bagging).
Commercial Operation: Phoenix deployed at Mark's retail stores (Canadian Tire subsidiary) performing real retail tasks — folding clothes, stocking shelves, bagging items.
Target Industries: Retail, logistics, light manufacturing, any task requiring fine manipulation.
Expected Production: Hundreds per year initially, with plans to scale to thousands once production-ready.
9. Engineered Arts
Country: United Kingdom | Founded: 2004 | Founder: Will Jackson
History: Engineered Arts is a British company based in Cornwall that has been building humanoid robots for two decades. Unlike most on this list, they focus on entertainment and human interaction rather than industrial labor.
Famous Models: Ameca (2021–present), Mesmer
Uniqueness: Ameca is widely considered the most human-like and expressive robot face ever built. Its ability to convey micro-expressions — raise eyebrows, smile, wink, react with genuine-looking emotions — has made it a viral internet sensation.
Pros: Unmatched facial expressiveness; reliable commercial product (sold and shipped); Android-style humanoid works perfectly for its niche; strong media presence (millions of YouTube views).
Cons: Ameca can't walk (stationary or wheeled base); not designed for industrial/manual labor; niche market (exhibitions, research, entertainment); expensive for what it does.
Key Breakthroughs: Ameca's facial expression engine (12+ distinct expressions); integration with GPT-4/LLMs for natural conversation; Mesmer's hyper-realistic silicone skin.
Commercial Operation: Ameca is commercially available today — sold to research labs, universities, museums, and tech companies worldwide. Each unit is hand-crafted.
Target Industries: Entertainment, museums, exhibitions, customer service, human-robot interaction research, education.
Expected Production: Tens to hundreds per year — boutique manufacturer focused on quality over quantity.
10. Apptronik (with NASA)
Country: USA | Founded: 2016 (UT Austin spin-off) | Key Partner: NASA
History: Apptronik spun out of the University of Texas at Austin's Human Centered Robotics Lab. The team originally worked on NASA's VALKYRIE (R5) humanoid robot for the DARPA Robotics Challenge. In 2023, they unveiled Apollo — a humanoid robot designed for commercial logistics, leveraging years of NASA-funded research.
Famous Models: Apollo (2023–present), Valkyrie/R5 heritage
Uniqueness: Apollo is designed from the ground up for manufacturability and modularity. Every component was designed to be easily mass-produced and swapped out. Their NASA partnership gives them access to world-class robotics IP.
Pros: NASA heritage and IP access; modular design for easy maintenance; partnerships with logistics companies; reasonable expected price (~$50K target); experienced engineering team from DARPA challenges.
Cons: Apollo less advanced in mobility than Boston Dynamics/Unitree; limited real-world deployments so far; smaller company with less funding; slower iteration cycle.
Key Breakthroughs: Modular actuator design for easy replacement; NASA Valkyrie-derived control algorithms; human-centered design (safety, easy interaction).
Commercial Operation: Apollo pilots underway with logistics and manufacturing partners. Deployed units remain limited.
Target Industries: Logistics, warehousing, manufacturing, eventually space exploration (NASA interest).
Expected Production: Hundreds to low thousands per year initially, scaling once manufacturing partnerships are secured.
Quick Comparison
- Best AI / Software: Figure AI (OpenAI-powered Helix model)
- Best Mobility / Agility: Boston Dynamics (Atlas — parkour, backflips)
- Best Price: Unitree Robotics (G1 at $16,000)
- Best Manufacturing Scale: Tesla (leveraging car production lines)
- Most Expressive: Engineered Arts (Ameca)
- Best Dexterity: Sanctuary AI (Phoenix — 20-DOF hands + haptic sensing)
- Most Commercial Deployments: Agility Robotics (Digit in Amazon/GXO/ERCO)
- Best Chinese Manufacturer: Unitree (aggressive pricing) / UBTECH (government-backed)
- Best for Home Use (Future): 1X Technologies (NEO — soft, safe, compliant)
- Best Space / Exploration: Apptronik (NASA Valkyrie heritage)
Bottom Line
The humanoid robot industry is experiencing its "iPhone moment" — the transition from research-lab curiosity to mass-market product. In 2023, the entire industry produced fewer than 100 humanoid robots. By 2026, that number is expected to hit thousands. By 2030, some analysts predict over 1 million humanoid robots could be deployed globally.
Key trends to watch:
- Cost collapse: Unitree's $16K G1 shows humanoid robot prices dropping fast. Within 5 years, sub-$10K humanoids may be common.
- AI integration: The real race is in AI — companies (Figure, 1X, Tesla) pairing great hardware with great AI will win.
- Manufacturing at scale: Winners will be companies that can mass-produce reliably. Tesla and Unitree have the best shot here.
- China vs USA: Chinese companies (Unitree, UBTECH) compete on price and manufacturing scale. US companies (Figure, Agility, Boston Dynamics) lead in AI and advanced capabilities.
- The home market: No humanoid is ready for home use yet. 1X NEO and Tesla Optimus are closest — both are still 3–5 years away from being household products.
The humanoid robot revolution isn't coming — it's already here. The question is not if these robots will transform our lives, but which company will lead the way.