Who are the leading humanoid robot makers?

By verified commercial deployment activity: US cohort includes Figure AI (Figure 02 at BMW, Figure 03 at Catalyst Reno), Apptronik (Apollo at Mercedes + GXO + Jabil), Agility Robotics (Digit at GXO Flowery Branch), 1X Technologies (NEO consumer), Tesla (Optimus internal-only), and Boston Dynamics (Atlas R&D). Chinese cohort includes Unitree (cohort price floor at $5,900 R1 / $13,500 G1), UBTech Walker (factory pilots), Xiaomi, XPeng, Fourier, and a deep long tail. Per DEPLOY's deployment bar, 8-12 makers clear the verified-pilot threshold.

Depends on use case; single-dimension ranking misleads. For consumer purchase: 1X NEO (only verified-consumer-available). For enterprise manufacturing: Figure 02 at BMW (30K-vehicle deployment depth) or Apptronik Apollo (3-customer breadth). For enterprise logistics: Agility Digit (100K-tote anchor at GXO Flowery Branch). For research and developer access: Unitree G1 ($13,500 base). For engineering credibility: Boston Dynamics Atlas. The "best" depends on which verification posture and deployment context matches the buyer's use case. See the best humanoid robot to buy for use-case-specific picks.

Per registry source-of-truth: Apptronik Apollo operates 3 Fortune-500 customer pilots (Mercedes-Benz, GXO Logistics, Jabil) for enterprise breadth leadership. Figure AI operates 2 generational anchors (Figure 02 at BMW Spartanburg with 30K-vehicle depth + Figure 03 at Catalyst Reno pilot). Agility Robotics Digit operates the cohort's single deepest customer relationship (100K-tote/year throughput at GXO Flowery Branch under multi-year RaaS contract). 1X NEO operates consumer commerce surface with pre-orders. Tesla Optimus operates ~300-500 internal-only units at Tesla facilities (classified as research per maker-facility rule, not commercial deployment).

Different vectors. Chinese cohort ships platforms at the most aggressive price points (Unitree G1 $13,500, R1 $5,900) and operates the broadest manufacturing-platform breadth (UBTech, Xiaomi, XPeng, Fourier, deep long tail). US cohort leads on enterprise customer verification depth (Figure at BMW 30K-vehicle anchor, Apollo at 3 Fortune-500 customers, Digit at GXO 100K-tote anchor) and consumer commerce surface (1X NEO). Different strategic differentiation; not a single ranking. See which humanoid robots are American vs Chinese for the geographic-origin disambiguation.

In the consumer-targeted segment: 1X NEO at $20,000 outright / $499/mo subscription (only verified-consumer-available humanoid). In the enterprise-deployed segment Optimus does not yet operate: Figure 02 at BMW, Apptronik Apollo at 3 Fortune-500 customers, Agility Digit at GXO. Per DEPLOY's verified-vs-claimed framework, Optimus competes at the trajectory-claim layer ($20-30K Musk-stated target) while NEO + Figure + Apollo + Digit compete at the verified-deployment layer. See Tesla Optimus vs Figure 03 vs 1X NEO for the 3-way cohort comparison.

Because the cohort operates at structurally different verification postures, deployment contexts, form factor optimizations, and labor-market interfaces. Ranking by single dimension (capability, pricing, deployment count) obscures the strategic differentiation that matters. Per DEPLOY's framework, four canonical dimensions matter: availability tier (consumer-available vs enterprise-deployed vs research-tools vs consumer-promised); capability tier (verified consumer-deployed vs verified enterprise-deployed vs research-and-demonstration); strategic posture (depth vs breadth vs consumer-direct); verification depth (per-deployment evidence vs per-cohort marketing claims). The serious distinction is between named pilot deployments and press-release-only claims.