Goldman Sachs Raises China Robotaxi Fleet Forecast to 3.1 Million by 2035, Initiates Coverage on Robotruck and Introduces Overseas Expansion Model
AI summary card
Goldman Sachs Raises China Robotaxi Fleet Forecast to 3.1 Million by 2035, Initiates Coverage on Robotruck and Introduces Overseas Expansion Model
The report significantly raises the China Robotaxi penetration rate forecast (reaching 36% by 2035), quantifies the overseas expansion path for the first time (overseas fleet share for WeRide/Pony AI/Baidu will rise to 43%/37%/38%), and introduces Robotruck forecasts for the first time (reaching 760,000 units by 2035).
- China Robotaxi fleet size forecast significantly raised: from 13k to 14k in 2026 (+195% YoY), and from 2.5m to 3.1m in 2035 (penetration rate reaching 36%)
- Overseas Robotaxi forecast introduced for the first time: overseas fleet share for WeRide/Pony AI/Baidu will rise from 29%/9%/7% in 2026 to 43%/37%/38% in 2035
- Robotruck sector covered for the first time: China's Robotruck fleet is expected to grow from 8k units in 2026 to 760,000 units in 2035 (accounting for 9% of heavy truck parc)
- Commercialization progress exceeds expectations: Pony AI, WeRide, and Baidu have achieved single-city unit economics break-even in multiple cities
- Key drivers include rapid cost reduction, accelerated software iteration, improved remote supervision efficiency, and continuously rising order density
Report interpretation
Overview
This report is Goldman Sachs' in-depth outlook and forecast for the global Robotaxi and emerging Robotruck markets. The core conclusion is that the commercialization process of Robotaxis in China is significantly accelerating, and fleet size forecasts have been systematically raised; meanwhile, the report builds a quantitative model for the overseas expansion of leading Chinese companies for the first time, and incorporates Robotruck, a long-term incremental sector, for the first time, comprehensively depicting the medium- to long-term development landscape of the autonomous driving mobility ecosystem.
Core views
The report believes that China's Robotaxi is transitioning from technology validation to scaled commercial deployment. The core drivers come from three aspects: first, rapid cost reduction, with lower vehicle and ADK costs combined with improved remote supervision efficiency; second, continuous operational optimization, with dual increases in order density (orders per day) and ASP per order, driving annual revenue per vehicle in Tier-1 cities to leap from USD 13,000 in 2025 to USD 38,000 in 2035; third, accelerated synergy between policy and commercialization, with multiple companies achieving single-city UE (unit economics) break-even. On this basis, the report significantly raises the China Robotaxi fleet forecast, increasing the 2035 penetration rate from the previous forecast of 31% to 36%, corresponding to a fleet size of 3.1 million vehicles. At the same time, the report proposes 'going global' as a key path to scale for the first time, forecasting that the overseas fleet share of WeRide, Pony AI, and Baidu will reach 43%, 37%, and 38% respectively by 2035. In addition, the report covers the Robotruck sector for the first time, dividing it into two major scenarios: 'closed roads' (ports, mines) and 'open roads' (trunk line logistics), forecasting that China's Robotruck fleet will reach 760,000 units by 2035 with a 9% penetration rate, where closed roads will be the main force in the near to medium term, and open roads will be the long-term growth engine.
Analysis framework
The report adopts a three-dimensional analysis framework of 'aggregate forecast + structural breakdown + commercialization validation'. First, based on historical deployment pace and clearer targets for 2026, it dynamically revises the aggregate forecast for China's Robotaxi fleet, and simultaneously updates the TAM (Total Addressable Market) value forecast (increasing from USD 56 million to USD 107 billion from 2025 to 2035). Second, it conducts a refined structural breakdown: expanding layer by layer by city tier (Tier-1/Tier-2/others), by operator (Pony AI/WeRide/Baidu/Didi, etc.), and by geographic distribution (domestic China/overseas markets), to identify growth elasticity and competitive landscape across various segments. Finally, it uses 'unit economics break-even' (UE break-even) as the core validation indicator for commercialization maturity, combined with the actual implementation progress of each company (such as Pony AI's break-even announcements in Guangzhou and Shenzhen, WeRide's in Abu Dhabi, and Baidu's in Wuhan), to substantiate the realistic foundation of the forecasts. For Robotruck, it adopts a 'scenario-first' analysis method, distinguishing regulatory difficulty and application breadth, modeling by scenario, and providing judgments on penetration rate inflection points.
Methodology notes
The core of the report revolves around the two-way validation of the supply capacity (fleet size, technology maturity, cost curve) and demand potential (mobility market penetration rate, logistics scenario adaptability) of Robotaxi/Robotruck
By analyzing vehicle supply growth, cost reduction trends, and demand-side changes such as user acceptance and increased order density, it comprehensively judges the industry's commercialization inflection point and market size ceiling.
Breaks down Robotaxi revenue into 'revenue per vehicle = order volume × ASP per order', and forecasts their growth paths respectively
The report explicitly points out that the growth in annual revenue per vehicle in Tier-1 cities mainly comes from the dual increase in daily average orders (from 20 orders in 2025 to 31 orders in 2035) and price per order (from USD 1.9 in 2025 to USD 3.4 in 2035), making the analysis more granular and verifiable.
Views Robotaxi as a systematic project, focusing on the coordinated evolution of the upstream (sensors, chips, algorithms), midstream (vehicle integration, fleet management), and downstream (mobility services, user payments)
The report mentions that cost reduction stems from 'lower vehicle and ADK costs' and 'improved remote supervision efficiency', which is precisely the result of the combined effect of upstream hardware cost reduction and midstream software efficiency improvement, reflecting the value chain transmission logic.
Asset mapping & comparison
Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).
- Pony AIOne of the core beneficiaries; the report focuses on tracking its domestic fleet expansion (target of 3,000 units in 2026) and overseas layout (overseas share of 37% in 2035)
- Strengths
- Has achieved single-city UE break-even for Gen-7 models in Guangzhou and Shenzhen domestically; rapid overseas fleet expansion (YoY growth of over 300% from 2025 to 2026E)
- Weaknesses
- Domestic market share (approximately 21% in 2026E) is lower than Baidu's (approximately 46%), facing fierce competition
- Comparison
- Compared to WeRide, Pony AI achieved single-city break-even earlier; compared to Baidu, it has a larger domestic scale but a slightly lower overseas share
- Risks
- Technology iteration risks, policy and regulatory uncertainties, geopolitical risks in overseas expansion
- WeRideOne of the core beneficiaries; the report focuses on tracking its global fleet targets (2,600 units in 2026, tens of thousands in 2030) and UE break-even validation in Abu Dhabi
- Strengths
- Aggressive global fleet expansion (target of 2,600 units in 2026), highest overseas share (reaching 43% in 2035E); Abu Dhabi fleet has validated UE break-even
- Weaknesses
- Domestic fleet scale (approximately 14% in 2026E) is lower than Baidu and Pony AI, localized operational capabilities remain to be verified
- Comparison
- Most aggressive overseas strategy, but relatively weak domestic foundation; a commercialization pioneer alongside Pony AI, but with slightly different profitability paths
- Risks
- Overseas operational compliance risks, domestic policy support not as strong as leading players
- BaiduOne of the core beneficiaries; the report focuses on tracking the rapid rollout of its Apollo Go in 20 domestic cities and UE break-even progress in Wuhan
- Strengths
- Largest domestic scale (fleet share of approximately 46% in 2026E), covering the most cities (approximately 20), with deep technological accumulation
- Weaknesses
- Overseas fleet share (38% in 2035E) is lower than WeRide's, with a relatively slower globalization pace
- Comparison
- Solid domestic leading position, but overseas expansion pace is slower than WeRide's; revenue per vehicle forecast in Tier-1 cities (USD 38k in 2035E) is on par with WeRide
- Risks
- Intensifying domestic market competition, lagging overseas business expansion, long commercialization profitability cycle
Key data
- China Robotaxi Fleet Size (2035E)3.1 million unitsRaised by 25% from the previous forecast (2.5 million units), with penetration reaching 36% (of total ride-hailing fleet)
- China Robotaxi TAM (2035E)USD 107 billionGrows approximately 190x from 2025 (USD 56 million), with a CAGR of over 60%
- Overseas Robotaxi Fleet Share (2035E)WeRide 43% / Pony AI 37% / Baidu 38%Significantly increased from 2026E (29%/9%/7%), reflecting that the overseas expansion strategy has become key to scaling
- China Robotruck Fleet Size (2035E)760,000 unitsAccounts for approximately 9% of the national heavy truck parc (8-9 million units), systematically covering this sector for the first time
Impact & implications
This report marks a new stage in Robotaxi industry research: shifting from proof of concept to proof of commercial scale. For related companies, it means the valuation logic will switch from 'technological leadership' to 'commercialization efficiency' (such as unit economics models, fleet density, and regional penetration rates); for investors, it is necessary to focus on leading players with scaled deployment capabilities, cost control advantages, and global operational experience. The introduction of Robotruck opens up long-term imagination for autonomous driving in the B2B logistics market, and its development pace will highly depend on the policy opening progress in closed scenarios (ports, mines) and the technology maturity on open roads (trunk line logistics).
Risks
- Policy and regulatory uncertainties: Licensing, road right opening, and safety liability determination for Robotaxi/Robotruck are still in a dynamic adjustment period
- Technology reliability risks: Large-scale commercialization places higher demands on system redundancy, extreme weather response, and long-tail scenario handling capabilities
- Capital expenditure pressure: Fleet expansion, R&D, and overseas deployment all require continuous massive investments, and the profitability cycle may be longer than expected
- Geopolitical risks: Overseas expansion of Chinese autonomous driving companies may face market access restrictions and data security reviews
What to watch
- The quantity and quality of 'unit economics break-even' (UE break-even) cities in each company's quarterly earnings reports
- Policy developments from the MIIT and MOT regarding the expansion of Robotaxi commercial operation pilot scopes
- Scaled deployment progress and customer contract signings for Robotruck in closed scenarios such as ports and mines
- Licensing approvals and cooperation progress for Chinese Robotaxi companies in key overseas markets (Middle East, Southeast Asia, Latin America)