Why Baidu Apollo matters
Among the handful of companies running fully driverless ride-hailing at commercial scale, Baidu's Apollo Go is the one that has paired Waymo-class operational depth with a genuinely global rollout. By April 2026 it had topped 22 million cumulative public rides, was operating across 27 cities, and had logged more than 330 million autonomous kilometers — over 220 million of them with no human in the driver's seat. In the first quarter of 2026 alone it delivered 3.2 million fully driverless rides, with weekly volumes peaking above 350,000.
What separates Apollo Go from Western peers is its posture toward expansion. Waymo has scaled deliberately, city by city, inside the United States; Apollo Go is exporting its fleet onto other operators' apps across the Middle East, Europe, and Asia. That makes it the clearest test of whether a Chinese autonomy stack, hardened on dense domestic streets, can travel — and whether robotaxis become a globally franchised utility rather than a set of national champions.
From an open platform to a robotaxi operator
Apollo began in April 2017 not as a service but as a platform: Baidu open-sourced its self-driving software and courted automakers and developers, betting it could become the Android of autonomous driving. The commercial pivot came in 2020, when Apollo Go opened public robotaxi rides in Beijing, and accelerated sharply in 2022, when Baidu won China's first permits to charge fares for fully driverless rides — with no safety operator at all — in Wuhan and Chongqing.
Those early driverless cities became the proving ground. Operating without a human backup at meaningful volume forced Baidu to solve the unglamorous problems of remote assistance, fleet logistics, and incident handling years before most rivals. The cumulative ride count compounding into the tens of millions is less a marketing figure than evidence of an operation that has run long enough to surface, and absorb, the long tail of edge cases that defines this business.
The RT6 and a cost-down strategy
The vehicle underwriting the scale-up is the sixth-generation Apollo RT6, unveiled in 2022 and now the workhorse of Apollo Go's international deployments. It is fully electric, purpose-built for autonomy rather than retrofitted, and carries more than 30 sensors feeding an onboard compute stack. Crucially, Baidu has pushed its bill of materials far below the cost of earlier modified-car fleets — the lever that turns a research program into a deployable business.
Cheap, replicable hardware is the strategic point. A robotaxi only earns when it is on the road, and a fleet only expands as fast as its unit cost allows. By engineering the RT6 to be manufactured at automotive scale, Baidu can flood new markets with vehicles at a capital intensity that competitors leaning on premium sensor suites struggle to match — and it can offer city partners and ride-hailing platforms a turnkey fleet rather than a science project.
Riding on Uber and Lyft instead of fighting them
Apollo Go's international model is distribution by partnership. Rather than build consumer demand from scratch in each new country, Baidu plugs its vehicles into the apps people already use. A multi-year global agreement with Uber, struck in 2025, commits to putting thousands of Apollo Go vehicles on Uber's platform; a parallel deal with Lyft targets Europe. Commercial driverless service on Uber went live in Dubai in early 2026, and on Abu Dhabi's Yas Island through local partner AutoGo, with a phased roll-out across the emirate.
Europe is the next frontier. On June 1, 2026, Apollo Go's AmiGo service — run with Swiss Post's PostBus — began open-road Level 4 trials across roughly 80 km² of eastern Switzerland under a special permit from the federal roads office, with regular fully driverless operations targeted for early 2027. London trials with both Uber and Lyft are slated for 2026. The bet is that, by being the fleet behind the world's biggest mobility apps, Apollo Go can scale internationally faster than any robotaxi operator trying to own the whole stack alone.
Economics: inside Baidu, edging toward profit
Apollo Go is not a standalone company but a division of Baidu (NASDAQ: BIDU; HKEX: 9888), and Baidu does not break out its revenue. That structure is both a cushion and a constraint: Apollo Go can fund a capital-hungry, long-horizon buildout from the parent's search-and-AI cash flows, but it competes for capital inside a company also pouring money into generative AI and cloud.
The financial signal investors watch is unit economics, and here the news has turned. On its first-quarter 2026 earnings call, management said Apollo Go had reached unit-economics breakeven in its largest China operating city despite very low domestic fares — and that overseas markets, where fares are higher, should carry stronger margins. With total rides up more than 120% year over year, the question is shifting from whether the technology works to whether the service can convert that growth into structural profit.
What to watch next
Three milestones will define the next year. First, whether the Swiss AmiGo trials shed their safety operators on schedule to reach genuinely driverless European operations in early 2027 — the first real proof that the China stack generalizes to Western regulatory and road environments. Second, whether the London launches with Uber and Lyft actually materialize amid the geopolitical scrutiny that Chinese-built autonomy now attracts abroad.
Third, and most important commercially, whether breakeven in one Chinese city becomes profitability across a 27-city, multi-continent footprint. Apollo Go has already answered the hardest engineering question — driverless rides at scale — affirmatively. The open questions are now political and financial: can it keep winning permits in markets wary of Chinese technology, and can the partnership-fleet model turn enormous ride growth into durable returns.
Sources
- Baidu Q1 2026 results (Apollo Go ride and footprint metrics)
- Baidu Investor Relations — Q1 2026 results and Apollo Go unit economics
- Apollo Go AmiGo Level 4 approval in Switzerland (PostBus)
- Baidu–Uber Apollo Go partnership, Dubai
- CNBC — Uber and Lyft to trial Apollo Go robotaxis in the UK
- Apollo Go — history and RT6 vehicle