Which humanoid robot makers are American, Chinese, or from other countries?

The geographic map of humanoid manufacturers

The humanoid robotics market is genuinely international in 2026, with major manufacturers operating from at least five countries. American consumers researching humanoid options often default-assume Chinese origin for unfamiliar brand names, which is sometimes correct and frequently wrong. This page is the geographic disambiguation reference for the full cohort.

The companion piece is Figure AI a Chinese company covers the Figure-specific disambiguation. This piece covers the cohort.

American humanoid manufacturers

The major US-based humanoid makers in 2026:

• Figure AI: Sunnyvale, California. Founded 2022 by Brett Adcock. American venture-funded with strategic backing from BMW Group, Microsoft, OpenAI, Nvidia, and others. Produces Figure 02 (BMW Spartanburg deployment) and Figure 03 (Catalyst Brands Reno pilot).
• Apptronik: Austin, Texas. Founded by Jeff Cardenas and Nicholas Paine; UT Austin spinout with NASA Valkyrie research heritage. Apollo platform deployed at Mercedes-Benz, GXO, and Jabil enterprise pilots.
• Tesla: Palo Alto, California (with Austin, Texas Gigafactory operations). Optimus humanoid program is part of Tesla's broader AI / robotics work; not yet a commercially-deployed humanoid platform.
• Boston Dynamics: Waltham, Massachusetts. Owned by Hyundai Motor Group since 2021. Spot quadruped is commercially deployed; Atlas humanoid is in research-to-commercial transition.
• Agility Robotics: Salem, Oregon (with Pittsburgh expansion). Digit platform deployed at GXO Logistics under the canonical commercial humanoid RaaS contract.

Chinese humanoid manufacturers

The Chinese humanoid cohort is substantial and ships at aggressive price points:

• Unitree Robotics: Hangzhou. Pricing-transparency leader with G1 ($13,500-$16,000) and R1 ($5,900) commercial pricing.
• AgiBot: Shanghai. Industrial and consumer humanoid development.
• UBTech Robotics: Shenzhen. Walker series with factory deployment pilots.
• Fourier Intelligence: Shanghai. GR-series humanoid platform.
• XPeng Robotics: Guangzhou. Iron humanoid; automotive-OEM heritage.
• Xiaomi Robotics: Beijing. CyberOne and successor humanoid programs.
• Additional Chinese makers in registry: Galbot (Beijing), LimX Dynamics (Shenzhen), Leju Robotics, Astribot, RobotEra, EngineAI, Booster Robotics, Noetix Robotics.

Other-geography humanoid manufacturers

• 1X Technologies: Norwegian-American structure. Originally Halodi Robotics (Moss, Norway). Subsequently operates a US footprint including the Hayward, California vertically-integrated humanoid factory opened April 2026. The NEO consumer humanoid is the company's flagship product; pre-orders open at $20,000 outright purchase or $499/month subscription (six-month minimum).
• Mentee Robotics: Israeli (Rehovot). Founded 2022 by Amnon Shashua (Mobileye founder). Acquired by Mobileye January 2026 for approximately $900 million; operates as Mobileye independent unit.
• Sanctuary AI: Canadian (Vancouver). Phoenix humanoid platform under development.
• Clone Robotics, Neura Robotics, PAL Robotics: European manufacturers across additional national footprints.
• Reflex Robotics, Skild AI, IntBot, Physical Intelligence: emerging US humanoid + brain-provider makers.

Why geographic origin matters less than the framework

Per DEPLOY's verified-vs-claimed framework, geographic origin is one editorial axis but not the determining axis for whether a manufacturer's claims are verified. The same framework discipline applies regardless of country:

• Verified-by-disclosure: 1X NEO's explicit teleoperation framing. The country of origin (Norwegian-American) doesn't change the framework's reading.
• Framing-without-disclosure: Tesla's We Robot 2024 framing of teleoperated Optimus as autonomous. The country (US) doesn't change the framework's reading.
• Verified pricing: Unitree G1 and R1 catalog pricing. The country (Chinese) doesn't change the framework's reading.
• Cap-flagged pricing: Apptronik Apollo, Figure 03 enterprise pricing. The country (US) doesn't change the framework's reading.

The differential verification postures across the cohort matter more than country of origin for buyers evaluating where humanoid commercial deployment actually lands. A Chinese maker with verified pricing operates at sharper verification than a US maker with claimed-pricing-no-commerce; the framework reads both honestly per their actual evidence anchors.

When geographic origin does matter

Three cases where country of origin is a meaningful editorial signal:

• Operational envelope: Chinese makers operating primarily in China face different regulatory frameworks, customer ecosystems, and supply-chain assumptions than US makers operating primarily in the US. Cross-geographic deployment scale is an additional verification surface.
• Supply-chain and trade considerations: tariffs, export controls, and trade-policy shifts shape what consumers and enterprise buyers can actually procure. The country shapes the practical buying surface.
• Capital structure and parent-company context: Boston Dynamics' Hyundai ownership; 1X's Norwegian-American structure; Mentee's Mobileye-Intel parent context. Institutional structure shapes operating priorities + commercial strategy + product roadmap.

Where to go for context

For consumer-evaluation context across the cohort: DEPLOY's per-model pricing pages for 1X NEO, Figure 03, Tesla Optimus, and the rolling-out Apptronik Apollo and Unitree G1 + R1 pricing pages.

For canonical institutional depth at the registry layer (per-maker founding history, leadership, deployments, source-depth verification), see the per-maker registry entries linked above.

For the framework DEPLOY applies to evaluating maker claims regardless of geographic origin, see how DEPLOY verifies capability claims and the canonical verified-vs-claimed methodology pillar. For methodology canonical references applicable to geographic-disambiguation framing: the 9-tier source-quality rubric (SEC + corporate IR + state-encouraged vs venture-funded source classification).