Humanoids Rise: Figure’s $1 Billion, Musk’s $1 Trillion, XPENG’s IRON

We discuss Figure’s $1 billion fundraise, XPENG’s (XPEV) humanoid launch, the humanoid market, and how Elon’s $1 trillion pay package fits into this.

With a $1 billion Series C Fundraise at a $39 billion valuation in September 2025, up from 15.19x from the $2.6 Series B in February 2024, Figure AI is now one of the world’s most valuable robotics companies. Figure introduced their Figure 3 humanoid about a month ago and have teased many videos demonstrating autonomous capabilities in industrial and household settings since. Their product is not available yet for mainstream/public purchase, but Figure currently has a target of 100 thousand annual production capacity by 2028.

If Figure can ramp up production and sell or rent 100,000 humanoids per annum, making ~$12 thousand per humanoid (rented, annual), that would be reaching a $1.2 billion revenue run-rate (added per annum). If you look at it this way, they are valued at ~32x 2028 potential revenue based on their goals. 

• Tesla (TSLA): $1.4 trillion market cap, 14.7x forward EV/sales 
• Hon Hai (2317), aka Foxconn: $110 billion market cap, 0.4x forward EV/sales
• Rockwell Automation (ROK): $43.8 billion market cap, 5.4x forward EV/sales
• FANUC (6954): $32.9 billion market cap, 5.1x forward EV/sales 
• XPENG (XPEV): $26 Billion Market Cap, 1.7x Forward EV/Sales

This is not a knock on Figure, as they have exhibited impressive progress. Rather, it’s a highlight of the opportunities existing in public markets for similar exposure, at discounted valuations, of companies that have already scaled production and operations globally. 

To be fair, the opportunity is quite large: Citi posits that the humanoid market could hit $7 trillion by 2050, with an estimate of nearly a billion humanoids. Now, if we are to assume that humanoids become mainstream, and if they are a hit, we could potentially see multiple per household (similar to cars, which average 1.8 per household in the U.S.). Now add the multitudes more in industrial/work settings, and this number would be drastically conservative, if calculated globally. 

For reference, let’s look at a related existing market, the automotive industry, to get a baseline understanding and comparison. Keep in mind, there is already a deep install base. That means that annual spend is going to be capped for automotive, across all class sizes. Meanwhile. there will be a growth spurt for humanoids, similar to how we are seeing AI datacenter buildout currently. 

• Light duty: about $1.2 to $1.4 trillion (60%–65% of total)
• Medium duty: about $200 to $300 billion (10%–15%)
• Heavy duty: about $210 to $275 billion (10%–15%)
• Combined total: about $2.1 to $2.8 trillion per year

Relatedly, Elon Musk just landed the opportunity to receive up to a $1 trillion dollar package upon achieving an enormously difficult set of goals: an $8.5 trillion dollar market-cap for Tesla, growing EBITDA from current ~$17 billion to $400 billion annual, and many operational milestones, such as a million Optimus robots delivered and a million robotaxis commercialized. He also must remain CEO for seven and a half years. As you’ll read ahead, this won’t be an easy feat. 

Winners may be fragmented, similar to the automotive industry. Expect multiple form factors, price points, regions, and end markets. There are 30‑plus humanoid startups and several public companies entering the field, including soon‑to‑IPO Unitree and listed firms with humanoid projects such as Hexagon (HEXA), UBTech (9880), Hyundai (by way of Boston Dynamics), 1X, Agility, Apptronik, and others.

Similar to the “mixture of experts” approach in AI, embedded intelligence will coordinate a mixture of robots that includes:

• Autonomous mobile robots for logistics, facilities, and basic home tasks where wheels win on cost and safety.
• Drones for inspection, scouting, light manipulation, and rapid small‑parcel logistics.
• Specialized robots in manufacturing, agriculture, healthcare, and energy with tight integration to industry tooling.

China-based XPENG was added to ROBO at the Q2 2025 rebalance. Founded in 2014, it is pursuing multi form‑factor robotics while scaling its core EV business. Revenue is projected to grow from about $11 billion in 2025 to about $20B in 2027.

At its AI Day in Guangzhou on November 5, 2025, XPENG unveiled the next‑gen humanoid robot named IRON. The demo appeared so human‑like that the team opened the outer covering on stage to show the spine‑like structure, bionic muscles, and articulated joints. Additionally, they claim to have implemented solid state batteries, which not only should have better energy density, but also are safer. Those factors are important as we start to bring these to our homes. 

For more news, information, and analysis, visit our Disruptive Technology Content Hub.