Apptronik Raises $935M at $5B Valuation for Apollo Robot

Inside the Round: Who's In and Why It Matters

The $520 million extension, co-led by B Capital, closed on the back of an oversubscribed $415 million initial Series A that Apptronik raised in 2025. Combined, the two tranches total more than $935 million at the Series A stage, one of the largest single-letter rounds in robotics history.

The investor composition tells the story that the press release doesn't. John Deere is the world's largest agricultural equipment manufacturer and has been quietly studying automation options for planting, harvesting, and logistics for years. A strategic check into Apptronik is a direct hedge on humanoid robots eventually reaching farms and farm-adjacent supply chains. AT&T Ventures suggests exploration of robots in telecom infrastructure: climbing cell towers, inspecting equipment in hard-to-reach locations, or managing data center operations. Qatar Investment Authority brings sovereign capital with a mandate to diversify Qatar's economy away from hydrocarbons, and robotics manufacturing is on that diversification list for several Gulf states.

Returning investors also deepened their positions. Mercedes-Benz has had Apollo in active pilots at its factory floor since March 2024, making the German automaker one of the earliest commercial customers of any humanoid robot company. Google's continued participation extends its DeepMind AI collaboration with Apptronik, where Gemini Robotics models run onboard Apollo.

Apollo: What the Robot Can Actually Do

Apollo stands 5 feet 8 inches tall (1.73 meters) and weighs 160 pounds (72.6 kilograms), placing it squarely in human scale. The design choice is deliberate: factories, warehouses, and offices are built for people, and a robot that matches human proportions can access the same workstations, shelving, and corridors without expensive facility modifications.

Payload is 25 kilograms (55 pounds), enough for most material handling tasks in manufacturing and logistics. Battery life reaches 4 hours per charge on a hot-swappable pack, meaning a facility running multiple battery sets can keep a robot operating for 20-plus hours in a shift cycle without downtime for charging. The locomotion system combines legs and optional wheeled configurations, letting Apollo cover ground quickly on flat surfaces while retaining the ability to navigate stairs and uneven terrain when needed.

On the AI side, Apollo runs Google DeepMind's Gemini Robotics models onboard. That's not just a marketing partnership: it's a deep integration where Apptronik co-develops training pipelines with DeepMind, using teleoperation data captured via MANUS haptic gloves to teach Apollo manipulation policies. The loop goes glove data to simulation to sim-to-real transfer to onboard policy, and the company has been running that pipeline at Mercedes-Benz factories for over two years.

AI-generated image

Apollo runs Google DeepMind's Gemini Robotics models onboard, trained via teleoperation data pipelines. AI-generated editorial image.

Apollo vs. Key Competitors

The Deployment Pipeline: Three Partners, Three Test Cases

Apptronik has three active commercial relationships that together cover the range of environments where Apollo is meant to operate. Each is at a different stage and tests a different set of capabilities.

None of these partners have published hard numbers on throughput, error rates, or ROI. That's standard practice at this stage: companies keep early performance data proprietary, and shared results tend to come later when programs have scaled past proof-of-concept and into operational commitments. The fact that Mercedes deepened its equity position in the Series A extension is the clearest proxy available for how the factory pilot is going.

Where the $935 Million Goes

Apptronik CEO Jeff Cardenas described the round as a "strong vote of confidence" in the mission. The company has been specific about how the capital will be deployed, and the plan is ambitious.

The new robot model announcement is worth watching. Apptronik hasn't disclosed what it will look like or what form factor it takes, but a company with $935 million in Series A funding and three active commercial programs building a second product in 2026 suggests the leadership team sees adjacent hardware opportunities that Apollo can't address alone. Some speculation points toward a more specialized arm or upper-body platform for tasks where legs add cost but not capability.

AI-generated image

GXO Logistics and Jabil pilots are testing Apollo in warehouse and assembly environments. AI-generated editorial image.

Competitive Context: The $1 Billion Club

Apptronik's $935 million Series A makes it one of the most heavily funded humanoid robotics companies not named Figure AI. Figure AI has raised over $1.8 billion and holds a $39 billion valuation, operating in a different tier. But below that outlier, the competitive picture is tighter than the headlines suggest.

The humanoid funding environment in early 2026 has been defined by a few dominant patterns. Chinese firms are raising fast (Galbot closed $350 million in March 2026, bringing total funding to $968 million), but their capital comes largely from state-adjacent sources (National IC Fund, Sinopec, CITIC affiliates), and their market access in the U.S. and Europe faces growing regulatory headwinds, including the Congressional national security hearing held March 18. American and European firms are raising from a broader base of strategic and institutional investors with more direct access to Western enterprise customers.

For Apptronik specifically, the Google relationship is a compounding advantage. The DeepMind Gemini Robotics integration gives Apollo access to frontier AI capabilities that most competitors have to build themselves or license at arm's length. As foundation models for robotics improve, companies with deep AI partnerships will likely see faster capability gains per dollar of R&D than those building AI stacks internally from scratch.

From NASA's Valkyrie to Factory Floors: The Apptronik Origin Story

Apptronik was founded in 2016 by researchers from the University of Texas at Austin's Human Centered Robotics Lab. The lab had a long history with DARPA-funded bipedal robot research, and several founders worked on NASA's Valkyrie space robot project, a humanoid designed for environments too dangerous for humans.

That background matters for understanding Apollo's design philosophy. NASA Valkyrie was built with safety and modularity as core requirements, not afterthoughts. Components had to be swappable by astronauts in spacesuits. Failure modes had to be graceful, not catastrophic. Those constraints shaped how the UT Austin team thought about robot hardware, and those instincts carried forward into Apollo.

The company spent its first several years building specialized robots for defense and research markets before pivoting to commercial humanoids. That period generated the kind of hardware engineering depth that shows up as fewer failure modes and more mature manufacturing processes, things that don't appear in spec sheets but matter a great deal when a robot is running 20 hours a day on a Mercedes factory floor.

Frequently Asked Questions

What This Round Signals for 2026

The Apptronik raise is not primarily a story about money. It's a story about which companies the market believes will still be standing in five years. A $935 million Series A from John Deere, AT&T, Qatar Investment Authority, Google, and Mercedes-Benz is not a speculative bet on unproven technology. These are organizations writing large checks with institutional review processes, fiduciary obligations, and reputational risk if they get it wrong.

That investor profile puts Apptronik in a category of its own among second-tier humanoid companies (behind Figure AI's outlier valuation). The company has the runway to scale Apollo production, develop a second platform, expand its AI collaboration with DeepMind, and convert active pilots into commercial contracts before competitors can close the gap.

The next critical milestones to watch: publication of any hard deployment metrics from Mercedes-Benz (throughput, uptime, error rates), the announcement of the new robot model, and any indication that GXO or Jabil are expanding beyond R&D trials into operational commitments. Those signals will tell you whether the funding was prescient or premature.