Apptronik
Pre-revenue hardware + robotics-as-a-service ambition: build and deploy Apollo humanoids for warehouse, logistics, and manufacturing labor, initially via paid commercial pilots (Mercedes-Benz, GXO) with a stated at-scale target price under $50K/unit; long-run model likely RaaS/leasing plus fleet software rather than one-time hardware sale. Apollo 3, teased Jun 2026, is positioned as the first true commercial product, with deployments targeted for 2027.
The Feb 2025 Series A and its Mar 2025 extension did not publicly disclose a headline valuation; the ~$1.8B figure is inferred from Bloomberg's '~3x the prior round' characterization of the Feb 2026 Series A-X (over $5.5B post-money). Series A-X valuationB uses Bloomberg's >$5.5B post-money; CNBC/TechCrunch cite a slightly lower ~$5-5.3B range.
Earnings, margins, COGS & capex
Apptronik is a pre-revenue-scale, venture-funded robotics company. It has raised more than $935M in Series A capital across an initial $350M (Feb 2025), an extension to ~$415M, and a $520M Series A-X (Feb 2026), for nearly $1B total, at a ~$5B+ valuation (~3x the prior round). There is no disclosed revenue, margin, or profit; the company is in a capital-intensive ramp phase, spending on Apollo production, a next-gen commercial robot (Apollo 3, targeted 2027), and a dedicated data-collection factory. The financial thesis rests on partner-validated pilots (Mercedes-Benz, GXO) and the Google DeepMind AI partnership, not on reported financials.
Income statement — where each revenue dollar goes
% of revenueOf every $1 of revenue, ~50¢ is cost of goods and ~50¢ operating expense, leaving ~0¢ of operating profit.
Revenue trend
Margins
Negative-to-thin today; targeted improvement as volume drives unit cost toward the sub-$50K goal
Negative during scale-up
Cash-consumptive
COGS structure
Not disclosed. Bill of materials dominated by actuators, force/torque sensors, batteries, and onboard compute (NVIDIA). Early-unit COGS is understood to be well above the sub-$50K at-scale target; cost-down depends on actuator design, vertical integration, and volume.
Capex
Not quantified. Major disclosed capital projects: Apollo production-line ramp, an Austin expansion plus a new California office, the ~90,000 sq ft Robot Park physical-AI data factory (Austin, announced Jun 2026, plus a growing global network at partner/customer sites), and development of a next-generation commercial robot (Apollo 3), with commercial deployments targeted for 2027.
Latest earnings
Not applicable
No financial guidance. Operational milestones: next-gen commercial robot (Apollo 3) commercial deployments targeted for 2027; Robot Park data-factory build-out and global network expansion; expanded commercial/pilot deployments.
- Total capital raised
- ~$935M Series A (nearly $1B total)
- Post-money valuation
- ~$5B+ (Feb 2026; over $5.5B per Bloomberg)
- Apollo lift capacity
- 55 lb (25 kg)
- Apollo runtime
- ~4 hrs/battery pack; hot-swap in under 5 min for near-continuous (~22 hr/day) operation
- At-scale price target
- <$50,000/unit
Growth drivers
- Apollo commercial pilots converting to paid, scaled fleet deployments (Mercedes-Benz manufacturing, GXO logistics)
- Google DeepMind Gemini Robotics AI partnership improving Apollo autonomy/generalization
- Robot Park data factory generating proprietary manipulation training data to compound the AI advantage
- Manufacturing partnership with Jabil to scale Apollo production
- Secular labor-shortage / automation demand in warehousing, logistics, and auto assembly
- Cost-down toward the sub-$50K/unit at-scale target unlocking broader TAM
Bull & bear
Apptronik is a credible top-tier contender in the humanoid race with the rare combination of a validated industrial customer base, a best-in-class AI partner (Google DeepMind), and nearly $1B of aligned strategic capital - positioned to ride a multi-trillion-dollar labor-automation wave if Apollo converts pilots to fleets.
- Blue-chip commercial traction (Mercedes-Benz, GXO) is real demand, not vaporware - de-risks the go-to-market versus demo-only peers
- Google DeepMind's Gemini Robotics gives Apollo a frontier AI brain, arguably the single most important input to humanoid usefulness, without Apptronik having to build a foundation model
- Cap table (Google, Mercedes, Deere, AT&T, QIA) supplies capital, distribution, and end-market pull simultaneously
- ~3x valuation step-up (to ~$5B+) in one year signals strong investor conviction amid record robotics funding
- Robot Park data factory + Jabil manufacturing partnership attack the two hardest problems - training data and cost-at-scale - directly
- At ~$5B, Apptronik trades at a large discount to Figure AI (~$39B), leaving room to re-rate if execution keeps pace
Apptronik is a pre-revenue, cash-burning hardware company valued at ~$5B on the promise of a product category that has repeatedly missed its commercial timelines - its own first true product (Apollo 3) is not due until 2027 - competing against far-better-funded rivals and commoditizing Chinese entrants, with its AI edge rented from a partner that could walk.
- No disclosed revenue and unproven unit economics - the entire valuation is a bet on a future that is years and billions of dollars away, with the first commercial robot slated for 2027
- Out-gunned on capital and stated production: Figure AI (~$39B), Tesla Optimus (mass-production ambitions, sub-$30K target), and Boston Dynamics (~$20B, Hyundai-owned) dwarf Apptronik's resources
- Chinese makers (Unitree, UBTech) are already shipping units at low prices, threatening to commoditize the hardware layer before Apptronik reaches scale
- AI dependency: Gemini Robotics is Google's, not Apptronik's - Google could favor a competitor (it also powers Boston Dynamics' Atlas), raise its take, or build its own robot
- Humanoid duty-cycle reliability, dexterity, and safety at industrial scale remain unsolved industry-wide; pilots and prototypes are not fleets
- If robotics-funding enthusiasm cools before positive unit economics arrive, a capital-hungry Apptronik faces dilution or a down round
What it is worth
Private venture round + comparables (last-priced round vs humanoid peers; no public multiples possible without disclosed revenue)
<$3B (down round)
if pilots stall, Apollo 3 slips past 2027, cost-down disappoints, Chinese hardware commoditizes the market, or robotics-funding enthusiasm cools before positive unit economics arrive.
~$5B-$8B
valuation supported by continued pilot progress and strong cap table, with modest step-ups tied to deployment and Apollo 3 milestones; still pre-revenue-scale.
$10B-$20B+
if Apollo converts Mercedes/GXO pilots into committed fleets, Apollo 3 ships on its 2027 target, unit costs trend toward <$50K, and the Gemini Robotics edge holds - closing part of the gap to Figure AI as the category re-rates.
Last priced at ~$5B+ post-money (Series A-X, Feb 2026; CNBC/TechCrunch ~$5-5.3B, Bloomberg over $5.5B), ~3x the prior round. With no disclosed revenue, valuation rests on strategic option value, customer validation, and the AI partnership rather than fundamentals. Peer anchors: Figure AI ~$39B, Agility ~$2.1B (VC) / ~$2.5B (pending SPAC), Boston Dynamics ~$20B current (bullish IPO projections up to ~$88-103B).
SWOT
Strengths
- Deep-pocketed, strategically aligned cap table — Google/Alphabet, Mercedes-Benz, B Capital, PEAK6, John Deere, AT&T Ventures, Qatar Investment Authority
- Google DeepMind partnership gives Apollo access to Gemini Robotics foundation models - a differentiated AI brain most hardware-first rivals lack
- Real industrial partners (Mercedes-Benz plants, GXO warehouses) validating the use case, plus Jabil for manufacturing scale
- Strong technical pedigree — spun from UT Austin's Human Centered Robotics Lab; team behind NASA's Valkyrie; force-control actuator expertise for safe human-adjacent operation
- Purpose-built Robot Park data factory (plus a global network at partner/customer sites) to compound a real-world manipulation-data moat
Weaknesses
- No disclosed revenue and deeply negative economics; entirely dependent on continued venture funding
- Early-unit cost far above the sub-$50K at-scale target - unit economics unproven
- Behind better-capitalized rivals (Figure AI at ~$39B, Tesla Optimus, Boston Dynamics ~$20B) on funding and, in Tesla/Figure's case, on stated production volume
- First true commercial product (Apollo 3) not due until 2027 — current Apollo 2 is a prototype generation, so scaled revenue is still years out
- Reliance on Google for the AI stack creates a strategic dependency it does not control
Opportunities
- Massive labor-automation TAM in logistics, warehousing, and manufacturing amid structural labor shortages
- RaaS/leasing model could turn a capex-heavy robot into recurring, high-margin fleet revenue + software
- First-mover paid deployments with blue-chip industrials create reference customers and switching costs
- Physical-AI data flywheel - proprietary manipulation data improves the model, which wins more deployments
- Potential expansion beyond industrial into commercial/service and eventually household use
Threats
- Intense, better-funded competition (Tesla Optimus, Figure AI, Boston Dynamics/Hyundai, Agility) plus low-cost Chinese entrants (Unitree, UBTech) that could commoditize hardware
- Google DeepMind could favor a rival (it also powers Boston Dynamics' Atlas) or vertically integrate its own robot, eroding the AI edge
- Capital-intensity + long path to positive unit economics could trigger a down round if the robotics funding cycle cools
- Safety incident or high-profile pilot failure could set back enterprise adoption industry-wide
- Timeline risk — humanoids have a long history of over-promising and under-delivering on commercial readiness; Apollo's first commercial product slips to 2027
Moats, dependencies & bottlenecks
Moats
Powerful differentiator today, but it is a partnership Apptronik does not own or control, and Google also powers Boston Dynamics' Atlas - durability hinges on Google's continued favor.
Reference customers + integration/switching costs, but pilots are not yet locked long-term fleet contracts.
Real-world data flywheel can compound if deployments scale; unproven until fleets are live.
Deep hardware IP for safe human-adjacent operation; talent and design know-how, but replicable by well-funded rivals.
Investor-customers (Mercedes, Deere) provide pull, but do not guarantee exclusivity.
Dependencies
AI foundation model + strategic investor Apollo's autonomy leans on Gemini Robotics; loss or de-prioritization would blunt the core differentiator - and Google also supplies rival Boston Dynamics' Atlas.
Onboard + training compute GPU/edge-compute supply and cost; industry-wide dependency shared by all humanoid makers.
Contract manufacturing / production scale Scaling Apollo output depends on this manufacturing partnership executing.
Anchor customers / pilots Revenue path and validation concentrated in a few marquee pilots; GXO also runs rival Agility's Digit, so the account is contested. Slippage stalls the commercial narrative.
Pre-revenue burn requires ongoing raises; a robotics-funding downturn is existential.
Component supply chain BOM cost and availability gate the path to sub-$50K unit economics.
Advantages
- Frontier AI brain via Google DeepMind without building a foundation model in-house
- Validated blue-chip industrial customers (Mercedes-Benz, GXO)
- Strategically aligned, deep cap table (Google, Mercedes, Deere, AT&T, QIA, B Capital, PEAK6)
- Force-control actuator IP for safe human-adjacent operation; NASA Valkyrie / UT Austin pedigree
- Dedicated data-factory (Robot Park) plus a global network to build a proprietary data flywheel
- Jabil manufacturing partnership to attack cost and scale
Weaknesses
- No disclosed revenue; economics deeply negative and unproven
- Out-funded by Figure AI, Tesla, and Boston Dynamics
- Core AI capability is rented, not owned
- First commercial product (Apollo 3) not due until 2027; early-unit cost far above the at-scale target
- Exposed to low-cost Chinese hardware commoditization
- Long, uncertain path from pilot to profitable fleet
Bottlenecks
- Unit cost - bringing early-unit COGS down toward the sub-$50K at-scale target
- Manipulation/dexterity reliability at industrial duty cycles (grasping varied objects; near-continuous operation via battery swaps)
- Manufacturing scale-up - moving from prototype Apollo 2 units to a shippable Apollo 3 fleet in 2027
- Real-world training data volume to make the AI generalize across tasks and sites
- Converting paid pilots into large, committed multi-year fleet contracts
- Capital intensity and sustained burn versus better-funded rivals
Top signals & trends
Top signals
Strong investor conviction amid record robotics funding.
Broadens end-market pull (agriculture, telecom) and sovereign capital access.
Investing in the data + next-gen hardware flywheel.
AI brain keeps improving; also a dependency Google also extends to rival Atlas.
Commercial traction remains pilot/prototype-stage and opaque.
Risk of being out-scaled and out-capitalized in a winner-take-most race.
Trends
Falling barrier to robot generalization; directly benefits Apollo via Gemini Robotics.
Demand tailwind for automation of dull/physical tasks.
Access to capital, but also inflates competition and valuations.
Hardware commoditization pressure on price and margin.
Winners convert pilots to fleets in 2026-2027; laggards stall - execution-defining phase.
Emerging public comps and a potential future exit path / valuation benchmark for Apptronik.
Ecosystem & competitor graph
Suppliers feed the company; customers pull from it. Line thickness shows the strength of each tie (supply-chain dependency, customer earnings contribution). Hover to isolate a tie.
Onboard + training compute (Jetson/GPU), Isaac/GR00T robotics stack.
Gemini Robotics foundation models - Apollo's AI brain (also investor).
Contract manufacturing partner to scale Apollo production.
BOM components; largely private specialist suppliers.
Commercial pilot of Apollo in manufacturing/assembly facilities (also investor).
Multi-phase warehouse/logistics deployment of Apollo (world's largest pure-play contract logistics); also runs rival Agility's Digit.
Best-funded pure-play humanoid (~$39B post-money, Series C exceeding $1B, Sep 2025); building its own Helix AI; targeting BMW and logistics. Private.
Mass-production ambitions, sub-$30K target price, in-house AI + manufacturing scale; internal deployment first.
Electric Atlas (Gemini-powered) shown CES 2026; Hyundai took 100% control Jun 2026 (bought out SoftBank's final stake for $325M). Currently valued ~$20B; bullish IPO projections up to ~$88-103B; Nasdaq listing earliest ~2027. HYMTF is the Hyundai-parent OTC proxy - Boston Dynamics itself is unlisted.
Digit humanoid focused on warehouse logistics ($30/hr RaaS); raised ~$400M Series C at ~$2.1B, now going public via a ~$2.5B SPAC merger with Churchill Capital, expected to trade on Nasdaq as AGLT by end-2026. Direct GXO/warehouse overlap.
OpenAI-backed (NEO humanoid); targeting home/commercial. Private.
Phoenix humanoid with dexterous manipulation focus. Private.
Hong Kong-listed Chinese humanoid maker (Walker series) shipping to auto plants; low-cost. Context only, not a buy/own call.
Chinese maker shipping low-cost humanoids/quadrupeds at high volume; commoditization threat. Private; context only, not a buy/own call.