31 Humanoid Robotics Companies Raise $8.2B — and Can't Hire Fast Enough

Skild AI closed a $1.4B Series C in January 2026, tripling its valuation from $4.5B to over $14B in seven months. The round was not an outlier — it was the predictable consequence of $8.2 billion-plus pouring into 31 humanoid-robotics companies in just over a year, and the handful of people who can bridge the gap between a prototype that works in the lab and a robot that works in a BMW plant.

The Bottleneck Nobody Is Talking About

The funding wave is staggering by any historical measure: 101 deals, eight-figure medians, valuations tripling in months. Figure AI hit a $39B valuation in September 2025, a 15x increase in 18 months. Apptronik raised a $520M Series A extension in February 2026 at a $5.5B valuation. NEURA Robotics pulled in roughly $1.2B in its Series C, reaching $4.6B.

But the real story is not the capital. It is what that capital is trying to buy next — and cannot find fast enough.

The bottleneck between prototype and production in humanoid robotics is no longer actuators, capital, or even foundation models. It is the small, non-obvious pool of engineers who can build physics-accurate digital twins and simulation environments capable of training generalist robot policies at scale. These engineers — who combine deep learning, kinematics, and real-time physics simulation — are the scarce, expensive, and largely silent center of the humanoid-robotics gold rush.

Neither the job market nor the mainstream conversation has caught up to how dramatically this single talent pool now shapes which companies ship product and which burn through funding waiting.

The Money Has Already Voted: Humanoid Robotics Is No Longer a Science Project

The sheer volume and velocity of capital have moved the sector from speculative R&D into early commercial deployment — and that shift changes which skills matter most.

Between July 2025 and June 2026, 20 disclosed deals across 15 unique companies raised $4.40B in pure-play humanoid robotics. The median round size was $144.85M; the average was $220.21M. Sixteen of 20 rounds exceeded $100M. Nineteen of 20 were follow-on financings — companies that already proved enough to raise again, not first-time bets.

Globally, robotics startups raised $13.8B in 2025, up from $7.8B in 2024. 2026 is on pace to exceed 2025. The humanoid sector alone crossed $500M in global sales revenue for the first time in 2025, and Goldman Sachs projects the broader humanoid robotics market will reach $38B by 2035.

The top 10 humanoid companies have captured nearly 80% of all capital raised in the category since 2022. Asia-Pacific accounted for 16 of 20 disclosed deals and $2.82B — 63.99% of capital — in the July 2025–June 2026 period. China's Ministry of Industry and Information Technology issued guidelines targeting mass production of humanoid robots by 2025, and a National Humanoid Robot Innovation Center opened in Beijing in 2024.

This funding environment has crossed a threshold where "build it and hope" is now "ship it or die" — which is what makes the simulation bottleneck so consequential.

From Lab to Line: Why the Real World Is the Hardest Problem Left

The defining engineering challenge in 2026 is no longer getting a robot to do a task once in a controlled setting. It is getting that robot to do it thousands of times, in varied conditions, without failure. Simulation is the only path to that scale.

Figure 02 logged over 1,250 hours at BMW Group Plant Spartanburg over 11 months, loading 90,000+ parts and contributing to production of more than 30,000 BMW X3 vehicles. Figure's BotQ factory scaled 24x in 120 days — from 1 robot per day to 1 robot per hour — reaching 55 robots per week by May 2026. Figure 03 launched in October 2025, redesigned for high-volume manufacturing and home environments, and was named a TIME Best Invention of 2025.

Agility Robotics' Digit has moved over 100,000 totes in real warehouse operations with GXO Logistics and Amazon. The company operates a 70,000-square-foot manufacturing facility with capacity for 10,000 units per year. In February 2026, Agility announced a commercial agreement with Toyota Motor Manufacturing Canada; in December 2025, it signed with Mercado Libre.

Apptronik's Apollo is deployed with Mercedes-Benz, GXO Logistics, and Jabil. Google chose Apollo as a flagship hardware platform for its Gemini Robotics project. Apptronik is targeting commercial production in 2026 with $1B in projected demand from current commitments.

1X Technologies' NEO first-year production capacity sold out within five days of pre-orders opening in October 2025. The company opened a 58,000-square-foot factory in Hayward, California, targeting 10,000 NEO robots in its first year and 100,000 units by the end of 2027.

The companies closest to real revenue are the ones most dependent on simulation and digital-twin infrastructure — because physical iteration is too slow and too expensive at this stage.

The Hidden Stack: Why Simulation and Digital Twins Are the New Moat

The companies pulling ahead are not just better at mechanical design or foundation models. They are better at building the simulation and digital-twin layers that turn a model into a reliable, deployable system. This layer is now the strategic moat.

Skild AI's "Skild Brain" can control robots it has never trained on, adapting in real time to extreme changes. The system uses four data sources: large-scale simulation, internet videos, teleoperation, and real-world deployments. Skild grew from zero to roughly $30M in revenue in a few months during 2025.

NVIDIA Isaac Sim, Isaac Lab, and GR00T are the default simulation and training stack for humanoid and generalist manipulation in 2026. MuJoCo remains the controls and dynamics simulator of record for whole-body MPC research. Drake, built by TRI and MIT, is a key manipulation simulator. ROS2 Jazzy and Humble is the dominant middleware.

NEURA Robotics and AWS announced a partnership in April 2026 to scale the Neuraverse platform. NEURA and the Technical University of Munich launched Europe's largest scientific training center for physical AI in March 2026. Boston Dynamics retired its hydraulic Atlas in 2024 and revealed a fully electric next-generation Atlas designed for commercial deployment.

The technical layer is not a supporting tool. It is the core competitive differentiator — and the companies that own it will define the sector.

The Talent Pool Is Tiny, Nameable, and Being Poached in the Open

The number of engineers who can credibly combine perception, deep learning, and kinematics for generalist manipulation on a real bipedal platform is small enough that recruiters can usually name the company and project a candidate worked on. That scarcity is driving compensation into territory that rivals or exceeds AI research labs.

KORE1 estimates there are roughly 2,000 engineers in the US who can combine perception, deep learning, and kinematics for generalist manipulation policy work. Senior humanoid experience commands a 15–35% premium over published compensation bands. The Bay Area pays 12–20% above national bands for senior engineers with humanoid or VLA experience.

Austin has become a Tier-1 robotics market driven by Apptronik and Tesla. Pittsburgh is now at near-parity with the West Coast for senior ROS and perception roles due to aggressive recruiting from humanoid companies. KORE1 reports a 92% twelve-month direct-hire retention rate on robotics placements — when you find the right person, you do not let go.

Well-scoped robotics searches close in 6 to 12 weeks. Learning or VLA specialist roles take 8 to 16 weeks. The IEEE Robotics and Automation Society member directory, CMU Robotics Institute alumni, TRI/Berkeley AI Research/MIT CSAIL alumni, and ICRA/RSS/IROS/CoRL conference attendance lists are the key sourcing channels. Everyone knows everyone. The pool is not growing fast enough.

The Compensation Curve Has Bent: $300K Is the New Floor for the Right Profile

For the specific intersection of simulation, digital twins, and whole-body control on a real humanoid platform, compensation has decoupled from broader robotics benchmarks — and the gap is widening, not closing.

Here is what the top humanoid companies are actually paying on the West Coast in 2026, per DeepRec.ai's US Robotics Salary Report:

For comparison, ZipRecruiter reports the average Robotic Simulation Engineer salary at $105,605 per year. Salary.com puts the average Simulation Engineer at $80,872. PayScale lists the average Robotics Engineer at $99,679. Glassdoor's median Robotics Engineer total pay is $142,400. Built In reports median base plus equity at $156,800. Levels.fyi's median at venture-backed and FAANG-adjacent companies is $232,000.

KORE1's placement data shows mid-level robotics engineers with 3 to 6 years of experience earning $150K to $205K base, and senior engineers with 6 to 10 years earning $205K to $300K. Humanoid and foundation-model specialists earn $280K to $475K in total comp. Senior Robot Learning and VLA Specialists with 6 to 10 years earn $250K to $345K base; Principal and Staff level earn $360K to $475K.

The headline $300K-plus figure is not an exaggeration for the right profile. It is the market.

The Hiring War Is Going Public — and Crossing Industries

The bidding war has spilled beyond pure-play robotics companies into Big Tech, automotive, aerospace, and industrial automation — all competing for the same narrow pool of simulation and digital-twin talent.

NVIDIA is hiring a Simulation Engineer for Industrial Physics and Robotics in Santa Clara. Microsoft is hiring a Member of Technical Staff for Robotics Simulation in Redmond. OpenAI is hiring a Simulation Realism Engineer in San Francisco. AMD is hiring a Principal Robotics Simulation Architect in Austin. Tesla is hiring a Virtual Commissioning Engineer in Fremont.

Hadrian is hiring a Factory Simulation Engineer for Omniverse in Los Angeles. SoftServe is hiring a Senior Omniverse Engineer and an Omniverse Architect, both remote US. Emerson is hiring a Projects Engineer for Digital Twin Simulation in Pittsburgh. Rockwell Automation is hiring Digital Twin Technical Consultants in Milwaukee and Detroit. XPENG is hiring a Senior Staff Digital World System Engineer in Santa Clara. Procore Technologies is hiring a Principal Software Engineer for Construction Digital Twin in Austin. SLB is hiring a Digital Twin Architect in Houston. Lockheed Martin is hiring a Simulation Systems Integration Engineer in Littleton. General Motors is hiring a Senior Manufacturing Engineer for Virtual Engineering in Warren.

Locus Robotics, GreyOrange, Symbotic, Vecna, and Amazon Robotics ROS2 platform teams are under-rated sources of mid-level robotics talent that humanoid companies are now actively recruiting from.

The war for this talent is not confined to the 31 tracked humanoid companies. It is spreading across the entire industrial and tech landscape, which will only intensify the scarcity.

What This Means for the Sector's Next 18 Months

The companies that secure and retain simulation and digital-twin talent in the next year will be the ones that cross into commercial production. The ones that don't will burn capital on physical iteration cycles they cannot afford.

Figure AI had 674 employees on LinkedIn as of May 2026. The company ran a 24/7 livestream of F.03 robots operating autonomously, reaching 191 consecutive hours and 238,000 packages by Day 9. Figure taught two F.03 robots to clean a room and make a bed in under 2 minutes, fully autonomously.

Sanctuary AI demonstrated zero-shot in-hand manipulation on a hydraulic hand, autonomously reorienting a lettered cube. Agility Robotics expanded its relationship with NVIDIA in March 2025 and continues to scale Digit deployments.

TrendForce projects Chinese humanoid robot output to grow up to 94% year on year in 2026, with Unitree and AgiBot expected to account for nearly 80% of global humanoid shipments. The competition is not just for talent — it is for the simulation infrastructure that determines how fast a company can iterate from prototype to production.

The simulation bottleneck is not a temporary hiring inconvenience. It is the structural constraint that will determine winners and losers as the sector moves from pilot programs to volume production.

The Quiet Engineers Who Will Decide the Humanoid Race

The humanoid-robotics industry has raised more money in 18 months than it did in the previous decade. It has signed deals with BMW, Mercedes-Benz, Amazon, Toyota, and GXO Logistics. It has crossed $500M in global sales revenue for the first time.

But the outcome of this race will not be decided by the size of the funding rounds, the elegance of the actuators, or even the sophistication of the foundation models.

It will be decided by a few thousand engineers who spend their days inside physics simulators, building digital twins of robots that do not yet exist at scale — and who, as of early 2026, can name their price.

The rest of the industry is just beginning to find out what they're worth.

Working in robotics? Zero G Talent tracks the openings: browse robotics jobs, the companies hiring, and the people building the field.