The people shaping what humanoid machines can become.
This is a provisional quarter-to-date research edition. Strategic Power measures a leader’s capacity to shape the humanoid field; it is separate from the Cyborg Index, which ranks robot offerings. Ties use competition ranking.
The ranking
18 shown
Co-founder and CEO · Tesla
Joint first: unusually integrated humanoid resources, agenda influence and options balance NVIDIA’s broader enabling reach.
Principal uncertainty
Tesla sustains robot funding and converts adjacent infrastructure into reliable manipulation without consuming its vehicle execution capacity.
Open full dossier
Founder and CEO · NVIDIA
Joint first: platform influence across competing humanoid makers balances less ownership of bodies and deployment environments.
Principal uncertainty
Robotics workloads continue to value NVIDIA’s integrated software enough to offset alternatives in cost, energy and supply resilience.
Open full dossier
CEO of Alphabet and Google · Alphabet
Third: Google’s mobilisable AI resources and explicit humanoid-model programme outweigh narrower programmes; hardware and adoption remain partner-dependent.
Principal uncertainty
Pichai can coordinate useful interfaces while preserving specialist teams and retaining the talent needed to advance robot intelligence.
Open full dossier
SVP, Google DeepMind; Google Chief AI Architect · Google DeepMind
Fourth: rare humanoid-model resources and operating oversight, bounded by Pichai’s allocation authority and OEM dependence.
Principal uncertainty
The new remit includes sufficient influence over robotics priorities, and partnerships supply the embodiment data and validation that general models lack.
Open full dossier
Executive Chair · Hyundai Motor Group
Joint fifth: exceptional industrial mobilisation balances less integrated model control; unrelated group robot and vehicle options are removed.
Principal uncertainty
Group leadership sustains coordination and capital through the slower work of industrial validation, maintenance and workforce integration.
Open full dossier
Chairman and CEO · XPeng
Joint fifth: rare vehicle-AI/chip/manufacturing transfer into IRON, with less mobilisation and humanoid ecosystem influence than the leaders above.
Principal uncertainty
The automotive and robot teams can share useful infrastructure without treating driving data as sufficient robot training or announced capacity as achieved output.
Open full dossier
Founder and CEO · Unitree
Seventh: humanoid hardware can shape developer choices; data capture and durable useful work remain less controlled.
Principal uncertainty
Unitree maintains an attractive development ecosystem and converts affordable, capable bodies into sustained use rather than a mainly demonstration-oriented market.
Open full dossierFounder, Chairman and CEO · AGIBOT
Joint eighth: balanced humanoid programme power, with less exceptional developer influence than Unitree’s estimated position.
Principal uncertainty
Product breadth yields reusable engineering and useful data rather than excessive customisation or predominantly promotional deployments.
Open full dossierChairman, Executive Director and CEO · UBTECH
Joint eighth: balanced humanoid engineering, mobilisation and commercial pathways; no dominant general-model or ecosystem claim earns 5.
Principal uncertainty
Humanoid sales translate into sustained useful operation, cash collection and reusable learning rather than primarily one-off customised projects.
Open full dossierFounder and CEO · Figure
Joint tenth: a focused body/model programme with exceptional options; repeat translation and platform dependence remain limited.
Principal uncertainty
Funding and future compute commitments convert into usable resources while Figure reduces intervention and service costs across several applications.
Open full dossier
Co-founder and CEO · Skild AI
Joint tenth: dedicated action models offer strong humanoid options; transfer and commercial claims remain material dependencies.
Principal uncertainty
Skild’s transferable robot-action layer can integrate with humanoid bodies, with technically useful and contractually permitted learning across embodiments.
Open full dossierCo-founder and CEO · Apptronik
Joint tenth: a focused humanoid body and major AI/manufacturing partners create options, while partner dependence constrains control.
Principal uncertainty
The partnerships supply dependable capacity and models while leaving Apptronik sufficient control over its product architecture and customer economics.
Open full dossier
Co-founder and CEO · Physical Intelligence
Joint tenth: reusable manipulation research can influence humanoid brains, but it is not a demonstrated complete humanoid controller.
Principal uncertainty
PI’s manipulation policies can be incorporated into humanoid systems with manageable hand, balance and whole-body adaptation costs.
Open full dossierPresident and CEO · Qualcomm
Joint fourteenth: a specific humanoid reference platform earns inclusion; general automotive/mobile reach adds no ecosystem bonus.
Principal uncertainty
OEMs adopt Qualcomm’s integrated stack in meaningful designs and the company sustains investment beyond demonstrations and reference hardware.
Open full dossier
Chair and CEO · AMD
Joint fourteenth: concrete humanoid computing contributions, with material but not pervasive humanoid developer dependence.
Principal uncertainty
AMD sustains software and embedded investment and robotics teams can adopt its platforms without disproportionate porting or integration costs.
Open full dossierCEO · Arm
Joint fourteenth: humanoid computing and named collaboration qualify; proposed coordination is not an adopted standard.
Principal uncertainty
Partners continue to value common architecture and development support, and proposed robotics coordination becomes useful implementation practice rather than solely marketing.
Open full dossier
CEO · Agility Robotics
Seventeenth: credible Digit translation, with a more concentrated logistics strategy and ecosystem reach.
Principal uncertainty
Existing customer work produces reusable operating lessons and financing supports expansion without disrupting service or product transitions.
Open full dossierFounder and CEO · 1X
Eighteenth: differentiated home-humanoid integration, but less repeat commercial translation and ecosystem influence than higher neighbours.
Principal uncertainty
Home deployment supplies useful, permitted learning data and assistance costs decline as reliable skills expand.
Open full dossierHow to read it
Strategic Power estimates how much capacity a leader can mobilise to shape humanoid outcomes. Each criterion is rated from 0–5 and weighted. Confidence is reported separately; missing evidence never silently becomes strength.
Coverage & limits
From a 61-person discovery pool, 18 leaders were fully assessed. Seventeen remain unresolved or incomplete; 26 were excluded by the stated scope.
The edition distinguishes established facts, attributed claims, analytical inferences, explicit assumptions and forward scenarios.
Revision r02 is a same-cutoff scope revision, not a movement edition. No arrows or inferred changes from r01 are shown.
Report ID: P1-2026Q3-asof-2026-09-13-r02
Read the machine-readable edition →