On this page
A Multi-Trillion Dollar Market
In June 2024, Elon Musk told Tesla shareholders that the Optimus humanoid robot could make the company worth $25 trillion, more than half the entire S&P...
In June 2024, Elon Musk told Tesla shareholders that the Optimus humanoid robot could make the company worth $25 trillion, more than half the entire S&P 500 at the time. Vinod Khosla, founder of Khosla Ventures, has predicted humanoid robots could eclipse the automotive industry by 2040.
So how big could this market actually become? And what makes analysts believe the opportunity is measured in trillions?
What Does "Biggest Product Ever" Actually Mean?#
To understand the scale of what's being predicted, we need context. The iPhone generated $201 billion in revenue for Apple in fiscal year 2024, representing about half of the company's total revenue. If you include the broader ecosystem of apps, services, and accessories, the smartphone market reaches approximately $1.2 to $1.3 trillion annually. This makes it one of the most successful consumer products in history.
The automotive industry is larger still. The global automotive market was valued at approximately $3.8 to $4.1 trillion in 2024, with vehicle sales reaching 84 to 91.6 million units annually. This includes not just car sales but the vast ecosystem of services, maintenance, insurance, and financing that surrounds vehicles.
Even larger sectors exist. Healthcare, estimated at around $9 trillion globally, and technology services at $5.5 trillion dwarf these single-product categories. But here's the critical distinction: these are entire industries, not single product lines.
When Musk claims Optimus could be "the biggest product ever," he's suggesting humanoids could surpass not just the iPhone or the car, but potentially entire industry categories. The mathematics of this claim rest on a simple observation: humanoid robots aren't selling devices or transportation. They're selling labor itself.
Selling Labor, Not Products#
The global labor market provides the essential benchmark. As of 2023, approximately 162 million Americans work, earning an average of $59,428 annually, totaling nearly $10 trillion in annual wages. Globally, with roughly 3.4 billion employed workers and average wages around $9,000, the labor market represents approximately $30 trillion annually, roughly 30% of global GDP (gross domestic product).
This is why humanoids are different from previous products. A household might buy one or two smartphones, perhaps two cars. But a factory might deploy hundreds of humanoid robots. A hospital might need dozens per shift. The unit economics aren't constrained by consumer purchasing patterns. They're driven by labor replacement calculations.
According to Goldman Sachs Research, the total addressable market for humanoid robots is projected to reach $38 billion by 2035, up more than sixfold from their previous projection of just $6 billion made only one year earlier. Their estimate for robot shipments increased fourfold to 1.4 million units over the same timeframe.
But even Goldman Sachs' revised figure may be conservative. ARK Invest forecasts that if humanoid robots achieve scale, they could generate approximately $24 to $26 trillion in revenues, split roughly equally between household and manufacturing applications. Morgan Stanley estimates the market could reach $5 trillion by 2050, with nearly 1 billion units deployed. Macquarie Research projects $139 billion by 2035 with 6.3 million units, growing to $1.7 trillion by 2050.
The spread comes down to one question: what percentage of the global labor market can humanoids actually capture?
The Economics of Adoption#
The mathematics of industrial adoption are surprisingly straightforward. ARK Invest calculates that if humanoid robots cost $16,000, they would only need to deliver slightly more than a 5% productivity gain over human workers to become economically viable. Unitree's R1 humanoid robot, launched in July 2025 at just $5,900, undercuts even that threshold.
Consider the hourly cost comparison. Macquarie analysts calculate that a robot priced at $100,000 operating 24/7 costs around $5.10 per hour, below the U.S. minimum wage and far below average manufacturing wages. As production scales and hardware costs drop to $20,000, the hourly cost falls to just $1.40.
The manufacturing sector provides the clearest near-term path. The U.S. manufacturing sector employs nearly 12 million people working approximately 23 billion hours annually for roughly $785 billion in total compensation. ARK Invest calculates that if humanoid robots were to work 16 hours per day (versus humans' typical 8-hour shifts), only 5.9 million robots would be needed to deliver the same manufacturing output.
Goldman Sachs' base case forecasts more than 250,000 humanoid robot shipments by 2030, with consumer robot sales ramping up to exceed 1 million units annually within a decade. Initially, deployment will focus on industrial applications: manufacturing, logistics, hazardous work. Goldman Sachs estimates humanoids could handle 5% to 15% of hazardous jobs and auto manufacturing roles.
But the true explosion comes when humanoids enter homes. Macquarie Research predicts home-based market demand will act as the tipping point, potentially propelling the humanoid category into a $3 trillion market by 2050. ARK Invest estimates a $13 trillion revenue opportunity in household tasks alone, automating activities like cleaning, cooking, and caregiving that are currently unpaid and unmeasured in GDP.
Why the Market Is Opening Now#
The convergence of AI, hardware, and economics that Part I described has a specific implication for market timing: the window for capturing this opportunity is opening now, not in some distant future.
Costs are dropping fast. Manufacturing costs have dropped from a range of $50,000 to $250,000 per unit to between $30,000 and $150,000 in just one year, far exceeding anticipated 15-20% annual declines. Unitree's $5,900 R1 represents a price point that seemed impossible two years ago. Tesla's Optimus is projected to cost under $20,000 when output reaches 1 million units annually.
Capital is flooding in. In September 2025, Figure AI announced it exceeded $1 billion in Series C funding at a $39 billion post-money valuation, representing a 15-fold increase from its $2.6 billion valuation just 18 months earlier. Investors include Microsoft, NVIDIA, and Jeff Bezos.
Deployment has begun. Tesla, Figure AI, Agility Robotics, and others are moving humanoids from laboratories into real workplaces, testing them in BMW plants, Amazon warehouses, and logistics centers. Agility Robotics' Digit is already earning revenue under a robot-as-a-service model.
Geographically, the race is global. Western companies lead in AI software models, while Asia is becoming the manufacturing hub for humanoid components due to its supply chain infrastructure and lower costs. China's humanoid robot market is projected to grow from approximately $378 million in 2024 to $10.3 billion by 2029.
What This Means#
The forecasts range from Goldman Sachs' $38 billion by 2035 to ARK Invest's $26 trillion. Even the low end describes a market larger than most industries today.
Whether humanoids become the largest product in history depends on challenges that haven't been solved yet. But the labor math, the falling costs, and the speed of AI improvement all point in the same direction.
In a market this size, the companies that move first could become enormous. One company in particular thinks it has the edge: Tesla.