Tesla and Figure vs China: Who's Building More Robots?
Tesla and Figure vs China: Who's Building More Robots?
16 hours ago•RoboStrategy•@robostrategy
YouTube15 min 8 sec
Watch on YouTube
Note: AI-generated summary based on third-party content. Not financial advice. Read more.
Quick Insights
  • Avoid chasing Unitree after its roughly 65% weekly surge; the company has warned of valuation risk, while profit fell about 17% as research costs rose.
  • For TSLA, treat Optimus capacity as a plan, not proof of commercial progress; look for verified production, customer deliveries, and robots reliably performing real tasks before investing on the robotics story.
  • Track Hyundai as a potential Atlas beneficiary, but wait for disclosed production volumes, costs, and commercial returns; full bookings for 2026 alone do not establish investment value.
  • Across the sector, prioritize evidence of recurring paid work—customer expansions, utilization, and task performance—over factory capacity or shipment totals.
Detailed Analysis

Tesla (TSLA)

  • Tesla says Optimus production has started on high-volume lines in Fremont, but it has not disclosed how many robots it has built.
    • A media report estimated production rose from dozens per week in Q2 to several hundred per week by August; the podcast treats this as unconfirmed.
  • Elon Musk said much of the equipment for a line designed to produce 1 million robots a year is installed, though reaching that rate would take at least a year.
  • Tesla plans to use robots internally to collect training data, with Musk saying it needs around 10,000 Optimus robots for that effort. Customer shipments are expected to start in about a year, initially for industrial and dangerous jobs.
  • Tesla identified the robot’s hand as its hardest remaining hardware problem. The podcast also notes that robots still need to prove they can perform real tasks at human-comparable speed.

Takeaways

  • Treat Tesla’s production capacity as a plan, not evidence of current output. Watch for independently verifiable production and customer-delivery numbers.
  • The key near-term test is whether Optimus can perform useful tasks reliably and at competitive speed—not whether Tesla can build a large production line.
  • The podcast’s suggested signals to watch are an unedited, convincing Optimus demonstration and progress on the hand and real-world task performance.

Figure

  • Figure’s CEO said the company built its 1,000th Figure 03 over the summer, up from just over 350 in April.
  • Figure reportedly keeps most robots in-house for research and testing, including a test home called “Black Box.” Its CEO said the company spent about $8.1 million on data in one week.
  • Figure 04 is design-locked, with parts already shipping. Figure’s CEO argued that developing robot intelligence matters more at this stage than simply building large numbers of robots.
  • Figure says it has more than $2 billion in cash and has arranged billions more in computing power for model training.
  • In a company-run 10-hour package-sorting test, Figure’s robot sorted 12,732 packages, compared with 12,924 for a person. The podcast cautions that this was a controlled test, not a typical warehouse floor.
  • The podcast host disclosed that Robot Strategy has an investment in Figure.

Takeaways

  • Figure’s strategy emphasizes training, data, and intelligence rather than maximizing near-term shipments. Track whether that approach translates into reliable deployments outside controlled tests.
  • The package-sorting result is promising but not conclusive; look for performance data from less controlled, customer-operated settings.
  • The host’s disclosed investment is relevant context when weighing the discussion of Figure.

Unitree Robotics

  • Unitree is a public company and one of China’s leading humanoid manufacturers. Its G1 robot was described as selling for roughly $13,000–$16,000.
  • The company says it shipped more than 5,500 humanoids in 2025. Its founder said it had built between 10,000 and 20,000 G1 robots since the model launched in 2024.
  • Unitree reported third-quarter revenue of about 665 million yuan, up roughly 70% year over year, while profit fell about 17% as research spending more than tripled.
  • Its shares rose about 65% in a week after a JP Morgan upgrade. Unitree then warned investors about valuation risks amid market optimism about physical AI.
  • Founder Wang Xingxing said a breakthrough in physical AI could be one to three years away, suggesting the technology still has significant development ahead.

Takeaways

  • The transcript describes strong sales growth, but also falling profit and rapidly rising research costs. Monitor whether growth can eventually support sustainable profitability.
  • The company’s own warning about valuation risk is a reason for caution after the sharp share-price rise. The transcript gives no stock ticker or price target.
  • Keep product shipments separate from proof of productive work: the podcast says it is still unclear how many robots are performing paid, useful work regularly.

Agibot

  • Agibot, based in Shanghai, reportedly built its 20,000th robot in September, after reaching 15,000 at the end of June.
  • A research firm cited in the podcast estimated Agibot shipped about 8,400 robots in the first half of 2026, making it the leading company by shipments in that count.
  • Its robots are being used in places including the Guangzhou subway system and a theme park. The company also demonstrated 110 production robots dancing in sync.
  • The podcast says Agibot proposed a “working” metric to measure robot activity, underscoring that shipment totals alone do not show how much productive work robots perform.

Takeaways

  • Agibot is a leading example of China’s manufacturing scale, but the more useful investment question is how many units are earning revenue through productive, recurring work.
  • Watch for utilization, customer deployments, and task-performance data—not just factory output or demonstrations.
  • The transcript does not provide public-market details or a valuation for Agibot.

Agility Robotics

  • Agility Robotics makes Digit and has not disclosed a total production count, so the podcast focuses on deployments instead.
  • Digit is at nine customer sites and has logged more than 65 hours on the job, according to a filing related to a planned public-market transaction.
  • At logistics company GXO, Digit has moved more than 100,000 totes. GXO has signed up for two additional warehouses starting early next year.
  • Agility’s reported revenue last year was about $1.8 million, indicating an early-stage business. Renting a Digit reportedly costs about $100,000 a year plus a setup fee; the podcast notes that this includes a different service and support offering than the lower-priced Unitree G1.

Takeaways

  • GXO’s additional warehouse commitments are a useful repeat-customer signal, but the small reported revenue shows commercialization is still at an early stage.
  • Track whether new deployments lead to recurring revenue and more customer expansions.
  • The transcript mentions a planned public-market transaction but gives no ticker, valuation, or specific investment terms.

Apptronik

  • Apptronik makes the Apollo robot and is running pilots with Mercedes-Benz, GXO, and Jabil.
  • The podcast says the company has raised close to $1 billion.
  • The host disclosed that Robot Strategy has an investment in Apptronik; the opening disclosure refers to the holding as “Uptronic.”

Takeaways

  • The pilots offer a way to assess whether Apollo can move from trials to recurring commercial use. Look for evidence of expanded deployments and measurable customer results.
  • The podcast provides no production count, revenue, valuation, or public ticker, so it does not establish a basis for evaluating the company’s investment price.
  • Consider the host’s disclosed investment when weighing the discussion.

Boston Dynamics and Hyundai

  • Boston Dynamics began building the production version of Atlas in January and says all Atlas units for 2026 are already spoken for.
  • The planned customers include Hyundai, which owns Boston Dynamics, and Google DeepMind.
  • Hyundai is planning for production of tens of thousands of robots over the next few years, according to the podcast. No current Atlas production count is provided.

Takeaways

  • Full 2026 bookings suggest demand for the initial production run, but the transcript does not say how many robots that represents or how much revenue it will generate.
  • For Hyundai, Atlas is a potential robotics growth opportunity, but investors would need further information on production volumes, costs, and commercial returns.
  • Google DeepMind is mentioned as a customer, not as a direct investment thesis for the robotics business.

1X Technologies

  • 1X, maker of the Neo home robot, opened a California factory that it says can produce 10,000 robots in its first year.
  • The company aims to begin customer deliveries by the end of the year, but the podcast gives no actual production count or delivery results.

Takeaways

  • The factory’s stated capacity is not the same as output or customer demand. Watch for confirmed deliveries and evidence that customers use the robots regularly.
  • The transcript does not provide a ticker, valuation, or operating financials.

Leju Robotics

  • Leju has 15 wheeled humanoid robots loading auto parts at a factory in Hangzhou and reportedly operating regular shifts.
  • The plant says the robots get the task right 98% of the time and have helped make more than 60,000 parts.
  • The reported loading cycle takes under 40 seconds, compared with about 15 seconds for a worker. Leju says data from the line is fed back into training.

Takeaways

  • Leju’s factory deployment is one of the transcript’s clearest examples of robots doing productive work, but the cycle-time gap shows that task economics still matter.
  • Track whether robot speed improves and whether the deployment expands beyond the current small group of units.
  • The transcript gives no public-market or valuation information for Leju.

Humanoid Robotics Industry and China–U.S. Competition

  • China’s industry ministry reportedly expects the country to build more than 100,000 humanoid robots this year, after producing more than 40,000 in the first half.
  • Different research firms use different shipment counts, but the podcast says all put China far ahead. One estimate says Chinese companies accounted for more than 97% of global shipments in the first half of 2026.
  • The podcast estimates U.S. companies built a low-thousands number of robots this year, compared with China’s expected 100,000. It says many U.S. companies are keeping robots in-house to gather data and train AI.
  • The transcript argues that higher production volume can lower component costs and put more robots in the field collecting data. It also stresses that volume alone does not ensure a robot can do useful work.
  • A cited analysis found the average price of a LiDAR sensor fell from about $530 to $260 in a year, illustrating potential cost declines in robot components.
  • The FCC added foreign-made mobile robots, including humanoids and robot dogs, to its “covered list.” The podcast says new Chinese robot models need Pentagon sign-off before they can be approved for sale in the U.S.; already-approved robots can remain.
  • The podcast says early applications are likely to focus on hard-to-fill or hazardous jobs, such as night shifts and heavy lifting. It suggests thinking in years, not months, while the technology develops.

Takeaways

  • The central investment question is whether companies can convert manufacturing scale into reliable, productive deployments. Compare robot task times with human task times, not just production totals.
  • Monitor customer deliveries, repeat orders, utilization, and performance on real work shifts. The podcast identifies these as more meaningful evidence than demonstrations or factory capacity.
  • China’s cost and scale advantages may benefit its manufacturers, while U.S. companies are emphasizing data and intelligence. The transcript does not establish which strategy will win.
  • The FCC restriction is a specific U.S. market-access risk for new Chinese robot models; the podcast also notes that some U.S. startups may use Chinese-made robots underneath their own products.
  • The transcript gives no specific stock price targets or buy/sell recommendations.
Ask about this postAnswers are grounded in this post's content.
Video Description
China expects to build over 100,000 humanoid robots this year, so how does America stack up? The breakdown covers 97% of first-half shipments coming from China, AgiBot hitting its 20,000th robot, Figure's 1,000th Figure 03, and Tesla's Optimus line built for a million a year. Follow RoboStrategy: X/Twitter: https://x.com/RoboStrategy LinkedIn: https://www.linkedin.com/company/robostrategy/ RoboStrategy Newsletter: https://tinyurl.com/mrx5wx2r Scott Walter: https://x.com/GoingBallistic5 Herbert Ong: https://x.com/HerbertOng 0:00 China expects to build over 100,000 humanoid robots this year 1:59 Over 97% of first-half shipments came from Chinese companies 5:08 Figure built 1,000 Figure Threes 6:14 Optimus production has started in Fremont 10:28 FCC adds foreign-made humanoids to the covered list 12:41 Hands are still the hardest hardware problem Disclaimer: RoboStrategy, Inc. (Nasdaq: BOT) is a closed-end investment company registered under the Investment Company Act of 1940. Risks include the possible loss of up to 100% of invested capital, and there is no assurance the Fund's investment objective will be achieved. Shares are not bank guaranteed and are not insured by the FDIC. FP Strategies LLC (RoboStrategy Advisors) is the Fund's investment adviser. Views are those of the speaker and not necessarily those of the Fund or its Board. This discussion is provided for information and educational purposes only and does not constitute investment advice, a recommendation, or an offer to buy or sell any security. Forward-looking statements and opinions are based on current expectations, estimates, projections, and assumptions and are subject to change without notice. Actual outcomes and results may differ materially from those expressed or implied. Past performance is no guarantee of future results. Any references to prior investment experience, portfolio companies, or investment outcomes related to activities conducted outside of RoboStrategy are provided solely for background and informational purposes. Any referenced gains, returns, or investment outcomes may be unrealized, do not represent the performance of any RoboStrategy fund, and are not indicative of future results or future investment outcomes for RoboStrategy. Investing involves risk, including possible loss of principal. Shares of closed-end funds may trade at a premium or a discount to net asset value, and market price and net asset value will differ. Neither the Fund nor the Adviser can predict whether shares will trade at, below, or above net asset value. References to companies, technologies, or investments are illustrative only and should not be interpreted as investment recommendations. Investors should consider the Fund's investment objectives, risks, charges, and expenses carefully before investing. This and other information is in the Fund's prospectus and reports, available at www.robostrategy.co/investor-center and at sec.gov. All views expressed are personal opinion as of date of recording and are subject to change without responsibility to update views.  No guarantee is given regarding the accuracy of information on this channel.  Neither host or guests can be held responsible for any direct or incidental loss incurred by applying any of the information offered. Author is an investor in RoboStrategy, Inc. (Nasdaq: BOT).
About RoboStrategy
RoboStrategy

RoboStrategy

By @robostrategy

RoboStrategy (NASDAQ: BOT) is the first fund dedicated to investing in the leading private robotics companies through a publicly listed investment company. Built to compound exposure. Leading robotics companies are still private, $BOT gives you access to them. Portfolio includes Figure AI, Apptronik, Standard Bots, Path Robotics, Dexmate, DYNA Robotics, and more.