2026-06-18 — views

Physical AI Manufacturing Scale — Waymo Zeekr Gen 6 vs Tesla Cybercab Gigafactory: AV Factory Economics

Waymo Gen 6 comes from Zeekr in China with 100-percent tariff risk; Tesla targets sub-$30K Cybercab at Gigafactory Texas with full vertical integration.

Article 156 in the Physical AI Benchmark Series — Physical AI Manufacturing Scale: Waymo Gen 6 Zeekr Production, Tesla Cybercab Gigafactory Ramp, and the Factory Economics of AV Fleet Production

Manufacturing scale is the bridge between AV technology and commercial viability. A robotaxi company can have the world’s best autonomy software, but if it cannot manufacture vehicles at cost and scale, the unit economics never close. Waymo’s Gen 6 vehicle is sourced from Zeekr — a Geely subsidiary manufacturing in China — then shipped to the United States for AV sensor stack installation. Tesla plans to manufacture the Cybercab at Gigafactory Texas at consumer-EV scale with full vertical integration, targeting a below-$30,000 manufacturing cost. This article is Article 156 in the Physical AI Benchmark Series. It benchmarks the two manufacturing strategies: their supply chain dependencies, cost structures, geopolitical risk profiles, and what factory economics mean for the AV unit economics race.

All figures labeled “(est.)” are derived from public disclosures, industry research, and analyst estimates rather than independently verified primary data.

Section 1 — Why Manufacturing Scale Matters for AV Economics

The economics of autonomous vehicle deployment are governed by fixed costs that must be amortized across a growing fleet. Every depot, every remote operations center, every HD map update, and every sensor calibration facility is a fixed cost that becomes cheaper per ride as the fleet grows. The manufacturing strategy — how vehicles are produced, at what cost, and at what volume — determines how fast a company can scale its fleet and how quickly it reaches the unit economics threshold for profitability.

Economic lever	Waymo situation	Tesla situation	Implication
Fixed cost amortization	$2–10M depot per city; remote ops center; HD mapping per city — all fixed costs spread across fleet	Gigafactory capex shared with consumer EV production; Cybercab manufacturing on existing lines	Tesla’s shared manufacturing infrastructure reduces AV-specific fixed cost burden
Vehicle unit cost at scale	Gen 6 unit cost not disclosed; Zeekr RT base chassis before AV stack; analysts estimate $30–50K per unit fully equipped (est.)	Cybercab target: below $30,000 manufacturing cost (Tesla stated)	Tesla has a stated cost target; Waymo has not disclosed per-unit economics
Fleet size to profitability	More vehicles → more rides → better fixed cost absorption → path to profitability	More vehicles → more FSD revenue plus ride revenue → faster profitability	Both need fleet scale; different revenue mix
Manufacturing bottleneck	Waymo depends on Zeekr production capacity plus import logistics from China plus US tariff exposure	Tesla controls its own manufacturing; Cybercab production constrained by Gigafactory capacity allocation	Tesla has supply chain control; Waymo has China dependency plus tariff risk
Vertical integration depth	Waymo: AV sensor stack plus software (deep); vehicle chassis (outsourced to Zeekr); battery (Zeekr)	Tesla: vehicle chassis, battery, FSD chip (HW4/HW5), software — nearly complete vertical	Tesla’s vertical integration enables faster cost reduction; Waymo’s outsourced chassis is a cost dependency
Scale economics path	Every new Waymo vehicle requires AV sensor stack installation (est. $5–15K per vehicle in AV hardware) on top of Zeekr base vehicle	Every Cybercab is designed from scratch as an AV — no retrofit of a consumer vehicle; AV hardware cost is designed in at the manufacturing level	Cybercab’s purpose-built approach may be cheaper at scale than retrofitting a consumer vehicle with AV hardware

The Manufacturing Economics Hypothesis

The core question in AV manufacturing economics is: at what fleet size and vehicle cost do the unit economics turn positive? The fixed costs — depot, mapping, remote operations — are largely independent of vehicle count above a city-level minimum. The variable costs — vehicle amortization, maintenance, cleaning, charging, insurance — scale directly with fleet size. A lower vehicle manufacturing cost compresses the payback period per vehicle, which accelerates the path to fleet-level profitability. Tesla’s below-$30,000 Cybercab target (if achieved) fundamentally changes the AV economics calculation compared to vehicles in the $50,000-plus range.

Section 2 — Waymo Gen 6: The Zeekr Partnership

Waymo’s Gen 6 vehicle represents a significant strategic shift from its Gen 5 platform. The Gen 5 used a Jaguar I-PACE — an expensive luxury EV not designed for AV fleet use, retrofitted with Waymo’s sensor suite. Gen 6 uses the Zeekr RT, a purpose-built AV-ready platform co-developed with Waymo’s operational requirements in mind. The partnership is rational given the constraints Waymo faced: no US OEM offered a purpose-built AV fleet platform at the required cost and volume.

Dimension	Detail	Notes
Vehicle base	Zeekr RT (Radically Transferable) — a purpose-built AV-ready platform from Zeekr (Geely Electric Automobile subsidiary, China)	Zeekr RT is designed specifically for AV fleet use: flat floor, no pedals, sliding doors, interior configured for passengers not driver
Why Zeekr	Geely/Zeekr can manufacture at scale; has existing EV manufacturing expertise; Zeekr RT was co-developed with Waymo’s AV requirements in mind	Waymo needed a manufacturing partner capable of producing thousands of vehicles per year; no US OEM offered a comparable purpose-built platform at the required cost
Manufacturing location	Zeekr vehicles manufactured in China; shipped to US for AV sensor stack installation by Waymo	Creates China supply chain dependency; adds logistics cost and lead time
US tariff exposure	Post-2024 US tariffs on Chinese EVs: 100-plus percent tariff on Chinese EV imports (Biden administration 2024, maintained under subsequent policy)	100-percent tariff would approximately double the import cost of each Zeekr RT chassis; severely impacts Waymo’s vehicle economics if applied to commercial AV fleet vehicles
Waymo tariff mitigation (est.)	Waymo has sought commercial AV fleet import exemptions; outcome uncertain as of mid-2026 (est.)	US commercial AV fleet vehicles may qualify for different tariff treatment than consumer EVs; regulatory pathway unclear
Gen 6 AV sensor stack installation	After Zeekr RT arrives in US, Waymo installs its proprietary sensor suite (lidar array, cameras, radar, compute module, cleaning systems)	This US-based installation step adds labor cost and creates a two-stage supply chain
Gen 6 production volume (est.)	Waymo has not disclosed Gen 6 production targets; fleet size growing from approx. 2,500 to est. 5,000–10,000 by 2027–2028 (est.)	Production ramp pace determines how fast Waymo can expand to new markets
Gen 6 cost vs Gen 5	Waymo disclosed “significant” cost reduction vs Gen 5 (Jaguar I-PACE); industry estimates 40–60 percent per-unit cost reduction (est.)	Gen 5 (Jaguar I-PACE) was an expensive luxury EV not designed for AV use; Gen 6 purpose-built design = structural cost improvement

The Two-Stage Supply Chain Problem

Waymo’s manufacturing process has two distinct stages separated by the Pacific Ocean. Stage one: Zeekr manufactures the RT chassis in China, including battery, motor, body, and base electronics. Stage two: Waymo receives the vehicle in the United States and installs its proprietary AV sensor suite — the lidar array, cameras, radar, compute module, and sensor cleaning systems. This two-stage process creates multiple cost and risk dependencies. The Zeekr stage is subject to Chinese manufacturing costs, Chinese labor costs, shipping logistics, port delays, and US customs and tariff policy. The US installation stage adds American labor costs. For every Gen 6 vehicle, Waymo must manage both stages successfully. If either stage is disrupted — by a tariff change, a port strike, or a Zeekr production issue — Waymo’s fleet growth plan is directly affected.

Section 3 — Tesla Cybercab: Purpose-Built AV at Consumer-EV Scale

The Tesla Cybercab is designed from first principles as an autonomous vehicle — not a consumer car with autonomy added on. It has no steering wheel, no accelerator pedal, no brake pedal. It is physically incapable of being driven by a human. This design decision reflects Tesla’s conviction that full driverless autonomy is achievable and imminent; it also means Cybercab inventory has zero value if driverless permits are not obtained.

Dimension	Detail	Notes
What is Cybercab	Two-seat, pedal-free, steering-wheel-free purpose-built AV; designed exclusively for robotaxi operation; not a consumer vehicle	No pedals/wheel = regulatory requirement to operate as AV only; cannot convert back to human-driven vehicle
Manufacturing location	Gigafactory Texas (Austin); will share manufacturing lines with Model Y (and eventually Cybertruck)	Same factory as Tesla’s existing production; incremental capex rather than greenfield
Price target	Below $30,000 manufacturing cost (Tesla stated, Musk multiple earnings calls)	At below $30K and $15 average ride fare, payback period potentially 3–4 years (est.) — positive unit economics
Production timeline (est.)	Tesla has stated Cybercab production begins 2026; volume production target for 2027 and beyond (est.)	Early 2026 production may be limited (hundreds of units); volume ramp is 2027 and later
Manufacturing scale target	Musk has referenced producing Cybercab at “high volume” eventually — implied millions per year as a consumer product eventually	If Cybercab production reaches 100K or more per year, it achieves consumer-EV manufacturing efficiency while being a purpose-built AV
Vertical integration advantage	Tesla manufactures: the chassis, the battery pack (4680 cells at Gigafactory Texas), the FSD inference chip (HW4/HW5), the software	No component dependency on any external supplier for the core AV stack; only commodity components (tires, glass, etc.) are external
Regulatory manufacturing risk	Cybercab has no steering wheel or pedals; requires separate regulatory approval in each state for use on public roads	If driverless approvals are slower than manufacturing ramp, Cybercab inventory accumulates without commercial revenue
Cost comparison to Waymo Gen 6 (est.)	Cybercab: below $30K target (Tesla stated) vs Waymo Gen 6: est. $30–50K per unit plus tariff exposure	If Tesla hits $30K target and Waymo faces 100-percent tariff on Zeekr chassis, Tesla could have 2–3x manufacturing cost advantage

The Gigafactory Shared-Infrastructure Advantage

Gigafactory Texas is already producing Model Y at scale. Adding Cybercab production to an existing factory with existing tooling, existing supply chains, and existing workforce is fundamentally different from building a new AV manufacturing facility from scratch. The capital expenditure required to add a new vehicle line to an existing Gigafactory is a fraction of the cost of a new greenfield factory. This shared-infrastructure advantage means that Tesla’s effective per-unit capital cost for Cybercab manufacturing is lower than the headline Gigafactory Texas investment. Waymo, by contrast, has no manufacturing infrastructure of its own — every vehicle must come from Zeekr, with all the supply chain dependencies that entails.

Section 4 — Supply Chain and Geopolitical Risk

The geopolitical risk profile of each company’s manufacturing strategy is dramatically different. Waymo’s Zeekr partnership creates deep exposure to US-China trade policy; Tesla’s Gigafactory Texas strategy creates domestic manufacturing insulation from that same risk.

Risk	Waymo exposure	Tesla exposure	Mitigation
China supply chain dependency	High — Zeekr RT manufactured in China; 100-plus percent EV tariff if applied; Geely is a Chinese state-linked company	Low for AV stack — Tesla manufactures key AV components in US; Model Y/Cybercab chassis in US (Gigafactory TX)	Tesla’s domestic manufacturing is a structural supply chain advantage vs Waymo’s China dependency
Semiconductor supply (compute chips)	Waymo custom compute module; relies on TSMC fabrication (Taiwan) for advanced node chips	Tesla HW4/HW5 also relies on TSMC (Samsung as secondary) for FSD inference chips	Both exposed to Taiwan semiconductor concentration risk; same for most tech companies
Rare earth materials (motors, batteries)	Zeekr batteries likely contain China-sourced rare earths	Tesla 4680 batteries: some China-sourced lithium/cobalt; domestic sourcing growing	Both have rare earth dependencies; Tesla has more domestic alternatives developing
Lidar supply chain (Waymo)	Waymo’s lidar sensors sourced from Luminar, Hesai, or proprietary; both US and China suppliers	N/A (Tesla camera-only)	Waymo’s lidar supply has both US and Chinese supplier options; diversified
US-China trade escalation scenario	Severe impact: 100-plus percent tariff makes Zeekr import economics untenable; Waymo would need alternative manufacturing partner (3–5 year transition est.)	Minimal direct impact on vehicle manufacturing; possible chip supply pressure	Tesla is structurally better positioned for US-China trade escalation
TSMC disruption (Taiwan) scenario	Would delay Waymo compute module production by 12–24 months (est.)	Would delay Tesla HW5 production by 12–24 months (est.)	Both equally exposed to Taiwan semiconductor risk; same for entire tech industry

The Tariff Scenario Analysis

The most acute near-term geopolitical risk for Waymo is the US tariff on Chinese EV imports. The Biden administration imposed 100-plus percent tariffs on Chinese EVs in 2024; subsequent administrations have maintained these tariffs as of mid-2026. If these tariffs apply fully to Waymo’s commercial AV fleet imports of Zeekr RT vehicles, the economics are severe. A Zeekr RT chassis estimated at $20–30K (est.) would cost $40–60K (est.) after a 100-percent tariff — before Waymo adds its AV sensor stack. At those economics, the below-$30,000 Cybercab target represents a potentially decisive manufacturing cost advantage for Tesla. Waymo’s primary mitigation is seeking a commercial AV fleet tariff exemption; the outcome of that regulatory effort is uncertain as of mid-2026 (est.).

Section 5 — Manufacturing Benchmark Scorecard

Dimension	Waymo Gen 6	Tesla Cybercab	Edge	2028 outlook
Vehicle unit cost target	Not disclosed; est. $30–50K plus tariff exposure (Zeekr RT plus AV stack)	Below $30,000 (Tesla stated)	Tesla (if target achieved)	Tesla decisive if Cybercab production starts and hits cost target
Manufacturing control	Outsourced to Zeekr; AV stack installed by Waymo in US	Fully vertical — chassis, battery, chip, software all Tesla	Tesla decisive	Tesla structural advantage in cost reduction pace
China supply chain exposure	High — Zeekr RT from China; 100-plus percent tariff risk	Low — Cybercab manufactured at Gigafactory Texas	Tesla decisive	Risk increases if US-China trade tensions escalate
Scale economics	Fleet of 2,500–10,000 vehicles (est.); never reaches consumer-EV manufacturing efficiency	Could reach 100K or more per year if Cybercab succeeds; consumer-EV manufacturing efficiency	Tesla (if scale achieved)	Tesla’s manufacturing scale target is transformative if achieved
Purpose-built AV design	Yes (Zeekr RT purpose-built for AV with Waymo)	Yes (Cybercab purpose-built AV, no pedals/wheel)	Even	Both benefit from purpose-built design vs Gen 5 retrofit
Regulatory-manufacturing alignment	Zeekr RT is designed for driverless; Waymo has driverless permits; manufacturing and regulatory aligned	Cybercab cannot be human-driven; if driverless permits delayed, manufactured vehicles cannot generate revenue	Waymo (regulatory-manufacturing alignment)	Tesla must ramp driverless permits in parallel with manufacturing

Overall Verdict

Tesla’s manufacturing strategy is structurally superior across the metrics that matter most for long-term AV unit economics: vertical integration, domestic production insulated from China supply chain risk, a stated below-$30,000 cost target, and the potential for consumer-EV manufacturing efficiency if the Cybercab ramp succeeds. Waymo’s Zeekr partnership is a pragmatic and rational solution to a real constraint — no US OEM offered a suitable purpose-built AV platform at the required cost and volume — but it creates China supply chain dependency, significant tariff exposure, and a two-stage production process that is inherently more complex than Tesla’s single-factory approach.

The decisive question is execution: can Tesla produce the Cybercab at the cost and scale it has stated, and can it obtain the driverless regulatory permits in parallel with manufacturing ramp? If yes, Tesla wins the manufacturing economics race decisively. If the Cybercab ramp is delayed, costs exceed $30,000, or driverless permit approvals lag manufacturing, Waymo’s more conservative supply chain approach — with established fleet operations and existing driverless permits — maintains competitive parity.

Manufacturing scale is the bridge between AV technology and commercial viability. As of mid-2026, Tesla is building a better bridge.

Note: All figures labeled “(est.)” are derived from public disclosures, industry research, analyst estimates, and reported data as of mid-2026. This article does not constitute investment advice or product recommendation.