Apptronik raises $520 million Series A-X at triple its earlier valuation, taking the round past $935 million as it readies a new Apollo
Austin humanoid maker Apptronik said on 11 February 2026 that it had closed a $520 million Series A-X extension priced at three times its Series A valuation, lifting total Series A funding above $935 million. John Deere, AT&T Ventures and Qatar Investment Authority joined returning backers B Capital, Google, Mercedes-Benz and PEAK6, and the company said the money will pay for Apollo production, training and data facilities and a new robot due in 2026.
Apptronik, the Austin, Texas, developer of the Apollo humanoid robot, said on 11 February 2026 that it had closed a $520 million extension to its Series A, which it calls Series A-X. The company said the new money brings its Series A to more than $935 million and its total capital raised to nearly $1 billion, making it one of the best funded humanoid start-ups outside China.
According to Apptronik, the extension was priced at three times the valuation of its original Series A. The company did not disclose the valuation itself. It put the first Series A tranche, raised in 2025, at $415 million. That figure is higher than the $403 million Apptronik reported on 18 March 2025, when it announced a $53 million add-on to its $350 million February 2025 raise; the company has not explained the $12 million difference, which may reflect later closings.
Returning investors named by Apptronik are B Capital, Google, Mercedes-Benz and PEAK6. New backers are AT&T Ventures, the venture arm of the US telecoms group; Deere & Company, the agricultural machinery maker that trades as John Deere; and Qatar Investment Authority, the Gulf state's sovereign wealth fund. Howard Morgan, chair and general partner at B Capital, which co-led the original round with Capital Factory, said the investor was deepening its support as Apptronik accelerates production and real-world deployment.
Apptronik said it will use the proceeds to ramp up production of Apollo, expand its global deployment network, build facilities for robot training and data collection, and develop a new robot that it plans to debut in 2026. Co-founder and chief executive Jeff Cardenas said in the release that the company is poised to unveil the newest version of Apollo. The company described its target markets as manufacturing, logistics, retail and healthcare, with the home as a later goal.
Apollo in its current form weighs 160 pounds, carries payloads of up to 55 pounds and runs about four hours per battery pack, according to Apptronik's product page. The packs are hot-swappable, so a robot can keep working while a spent battery is exchanged. Commercial users named by the company are Mercedes-Benz and GXO Logistics. Mercedes-Benz said in March 2025 that it had made a low double-digit-million-euro investment in Apptronik and was testing Apollo at its Berlin-Marienfelde Digital Factory Campus on tasks such as carrying parts to the line and first-pass quality checks.
Manufacturing is outsourced. Jabil, the US contract manufacturer, said on 25 February 2025 that it would build Apollo robots in its own facilities and pilot them inside its own production operations, with newly built units tested on Jabil lines before shipping to customers. Neither company has given a production volume or start date.
Apptronik spun out of the Human Centered Robotics Lab at the University of Texas at Austin in 2016. The company says it has built 15 robotic systems, including work on NASA's Valkyrie humanoid. Headcount has roughly doubled in a year, from more than 150 employees cited in February 2025 to nearly 300 now, by the company's own figures.
The robot's intelligence comes largely from Google. Google DeepMind said on 12 March 2025 that it was partnering with Apptronik to build the next generation of humanoid robots with Gemini 2.0, naming Apollo as an embodiment for its Gemini Robotics models. Google is also an equity investor in every Apptronik tranche disclosed since February 2025. That arrangement lets Apptronik concentrate on hardware, actuation and safety while a hyperscaler carries the cost of foundation model training.
The raise lands weeks after two markers set the competitive bar. Skild AI announced a $1.4 billion Series C at a $14 billion valuation on 14 January 2026, and Boston Dynamics used CES on 5 January to start building a production Atlas, with parent Hyundai targeting 30,000 robots a year. Figure AI's $1 billion Series C in September 2025 valued it at $39 billion. Against those numbers Apptronik's undisclosed valuation is the main gap in the picture.
- Apptronik
- US
- 2016
- Series A-X extension
- 520
- No data
- 2026-02-11
- Skild AI
- US
- 2023
- Series C
- 1400
- 14
- 2026-01-14
- Figure AI
- US
- No data
- Series C
- 1000
- 39
- 2025-09-16
- Agility Robotics
- US
- 2015
- Series B
- 150
- No data
- 2022-04-22
- 1X Technologies
- US
- 2014
- Series B
- 100
- No data
- 2024-01-13
Rounds as recorded in the ROBOTNESS company database from company announcements; not necessarily each company's most recent financing. Apptronik's valuation is undisclosed; the company says the Series A-X was priced at 3x its Series A valuation. Figures disclosed, not estimated; null = not disclosed.
As of Oct 1, 2026
Several questions remain open. Apptronik has not published revenue, units deployed or a price for Apollo, and has not said where the new training and data facilities will be or when Jabil-built units will ship. The arrival of John Deere and AT&T as investors suggests interest from agriculture and telecoms field work, but neither has announced a deployment.
The next checkpoints are the unveiling of the new Apollo, which Apptronik has promised for 2026, any announcement of the training and data collection sites, and evidence from Mercedes-Benz or GXO that pilots have moved to paid fleet deployments.
ROBOTNESS analysis
Apptronik has bought itself a place in the top tier of Western humanoid makers by renting its brain from Google and its factory from Jabil, and the 2026 robot launch will show whether that asset-light model can match vertically integrated rivals on cost and pace.
The evidence for the thesis sits in the investor list and the partner list. Google, Mercedes-Benz and Jabil each supply something Apptronik would otherwise have to build: frontier AI models, an automotive test floor and contract manufacturing at scale. A threefold step-up in valuation within roughly a year, which the company disclosed even while withholding the absolute number, shows that investors are paying for that network rather than for disclosed revenue.
The strongest counter-argument is that dependence cuts both ways. Figure AI develops its Helix model in-house and Tesla builds Optimus hardware and AI inside one company, so both control their learning loop end to end. If Gemini Robotics becomes available to many robot makers on equal terms, Apollo's AI advantage narrows to whatever data Apptronik collects in its own facilities.
Bull case. The new Apollo arrives in 2026 with Gemini Robotics built in, Jabil begins volume builds, and Mercedes-Benz, GXO and the new strategic investors turn pilots into multi-site orders. Apptronik's nearly $1 billion then funds a manufacturing ramp without needing a down round, and its valuation closes some of the distance to Figure.
Bear case. The redesign slips, Jabil volumes stay at pilot scale and the Google partnership proves non-exclusive. With roughly 300 staff and no disclosed revenue, Apptronik would then face a market where Skild AI, Figure and Chinese makers offering cheaper hardware set the price, and a further raise would come on harder terms.
Signals to watch:
- By 31 December 2026: the unveiling of the new Apollo, including payload, runtime and a target price.
- By 30 June 2026: the location and size of the robot training and data collection facilities Apptronik says it will build.
- Mercedes-Benz annual report for 2026, due in early 2027: whether Apollo use extends beyond the Berlin-Marienfelde test site.
- 3x the Series A valuation; absolute valuation not disclosed
- 11 February 2026, Austin, Texas
- Nearly 300 employees (more than 150 in February 2025)
- $520 million Series A-X extension
- 160 lb weight, 55 lb payload, about 4 hours per hot-swappable battery pack
- Apollo production ramp, global deployment, training and data collection facilities, new robot debuting in 2026
- AT&T Ventures, John Deere, Qatar Investment Authority
- More than $935 million (company says nearly $1 billion raised in total)
- B Capital, Google, Mercedes-Benz, PEAK6
Why it matters
The Series A-X puts Apptronik alongside Figure AI, which raised $1 billion in a Series C in September 2025, and Skild AI, which raised $1.4 billion in January 2026, among Western robot companies that have raised at least $500 million in a single round. It also confirms that strategic corporates rather than only venture funds are setting the pace: four of the seven named investors are operating companies or their venture arms (Google, Mercedes-Benz, AT&T, John Deere).
The threefold step-up is the most important disclosed number. Apptronik did not publish its valuation, but by stating the multiple it told the market that its price roughly tripled between early 2025 and early 2026, in line with the re-rating seen at Skild AI and Figure.
Rival analysis
Figure AI is the closest US analogue: a single humanoid platform aimed at automotive and logistics, with BMW as anchor customer. The difference is integration. Figure trains its own Helix model, while Apptronik relies on Google DeepMind's Gemini Robotics. Agility Robotics, with Digit in GXO warehouses, competes for the same logistics buyers.
Boston Dynamics, backed by Hyundai's plan for 30,000 robots a year announced at CES 2026, competes for automotive floor space and has a captive buyer Apptronik lacks. Tesla's Optimus is the vertically integrated extreme. Chinese makers such as Unitree, UBTech and AgiBot compete mainly on hardware price, which matters once Apollo moves from pilots to fleet sales.
Valuation context
Apptronik's absolute valuation is undisclosed, so any comparison is directional. Skild AI's $14 billion and Figure's $39 billion post-money values sit at the top of the private market; Apptronik's total Series A of more than $935 million is close to Figure's $1 billion single round and below Skild's $1.4 billion.
Investors are paying for partnerships and engineering pedigree more than revenue, since Apptronik has published no revenue or unit numbers. That leaves the valuation sensitive to the 2026 product launch: a delay would undercut the 3x step-up narrative quickly.
Supply-chain implications
Apptronik's model is asset-light. Jabil, which runs contract manufacturing for electronics and industrial clients worldwide, said in February 2025 that it would build Apollo in its own plants and use the robots on its own lines. That gives Apptronik access to global factories without building one, but it also means Jabil's priorities and pricing shape Apollo's cost curve.
The component chain for humanoids (precision reducers, frameless motors, batteries and compute modules) is shared with every rival, and Asian suppliers dominate several categories. Apptronik has not disclosed its actuator or compute suppliers, which leaves its exposure to tariffs and supply bottlenecks unclear.
Signals to watch
The first signal is the new Apollo, which Cardenas said the company is poised to unveil; Apptronik has committed to a 2026 debut. Specifications, a price and a launch customer would show how far the design has moved toward manufacturability.
The second is the training and data facilities. Where Apptronik builds them, and whether they sit with Mercedes-Benz, Jabil or the new strategic investors, will indicate which customers are closest to fleet orders. The third is any deployment announcement from John Deere or AT&T, which would open sectors beyond factories and warehouses.
Analyst view
Thesis: Apptronik's capital and partner network place it in the top group of Western humanoid makers, but it remains the least transparent of them on valuation, revenue and production volume. Confidence: medium.
Confidence is medium rather than high because the bull case rests on two milestones that have not happened (a new Apollo and Jabil volume production) and on a Google partnership whose exclusivity terms are not public. It is not low because the investor list, the threefold step-up and the doubling of headcount are all disclosed facts that point in the same direction.
Questions you should be asking
What is the post-money valuation, and how does the $415 million first-tranche figure reconcile with the $403 million the company reported in March 2025?
Is Apptronik's access to Gemini Robotics exclusive in any form, or can rival humanoid makers license the same models? What volumes has Jabil committed to, and in which countries will Apollo be built? Have Mercedes-Benz or GXO signed fleet orders beyond pilots?