China’s Robotaxi Giant Takes on Waymo—at Just One-Fifth the Cost
Baidu’s Apollo Go hits 250,000 weekly driverless rides, forcing investors to rethink the self-driving race
Baidu’s Apollo Go has quietly pulled even with Alphabet’s Waymo in the one metric that really counts: fully driverless rides per week. As of October 31, the Chinese tech powerhouse reports more than 250,000 autonomous trips weekly—matching Waymo’s output in the U.S.—but with cars that cost a fraction as much. While Waymo’s modified Jaguars run around $175,000 each, Baidu’s purpose-built vehicles cost closer to $30,000.
That number isn’t just a milestone—it’s a wake-up call. With 17 million global ride requests and over 240 million kilometers driven without a major injury, Baidu’s data suggests that the future of driverless mobility may be shifting from Silicon Valley to Shenzhen.
Benchmark investor Bill Gurley summed it up bluntly: “Apollo is neck and neck with Waymo. Key difference? The car’s way cheaper—$30k versus $175k.” That price gap is rewriting industry math.
And it’s not just about pride. At 250,000 rides across roughly 1,000 cars, each Apollo Go robotaxi is averaging about 250 trips per week—far more than Waymo’s 1,500-car U.S. fleet. In packed Chinese cities like Wuhan, with 12 million residents and dense commuting patterns, robotaxis operate almost nonstop. It’s an entirely different economic equation from sprawling American suburbs.
The Business Case: Scale, Cost, and the Road to Profit
For investors, this update turns hype into something tangible. In just four months, Baidu jumped from 169,000 to 250,000 weekly rides—a 48% increase that signals rapid growth and growing regulatory comfort across multiple Chinese cities.
Although Baidu hasn’t broken out detailed financials, analysts can piece together the economics. Its RT6 model, designed specifically for autonomous driving, takes advantage of China’s electric vehicle supply chain and Baidu’s own sensor systems to reach that low $30,000–$45,000 cost range. Even factoring in depreciation, remote monitoring, and mapping infrastructure, Baidu’s efficiency edge over Western competitors looks built to last.
Each vehicle runs nearly nonstop—20 to 25 minutes per trip, hundreds of rides per week. That constant use spreads out fixed costs fast. Combined with China’s lower energy, labor, and parking expenses, analysts believe city-level profitability could arrive within 12 to 24 months.
Safety performance adds another layer. Over 240 million kilometers driven without serious injury—including 140 million fully driverless—means fewer human interventions. Each mile makes the AI smarter and trims operating costs. The more the cars learn, the fewer humans Baidu needs watching over them.
Still, numbers can mislead. The widely cited “17 million cumulative orders” almost certainly includes uncompleted or canceled requests, so it’s more about scale than revenue. The 250,000 weekly driverless trips—that’s the figure that matters for comparison with Waymo’s paid rides.
So what’s that worth? If you model Waymo-level performance with far lower vehicle costs—and factor in Beijing’s strong backing for homegrown AI—a standalone valuation between $5 and $8 billion for Apollo Go looks reasonable. If Baidu confirms profitability data or announces international deals with revenue guarantees, that number could easily stretch past $10 billion.
This announcement also plays to two audiences. For Chinese regulators, it validates their bet on Baidu as the national champion. For global investors, it challenges the assumption that Alphabet owns the driverless future. Pony.ai CEO James Peng captured the moment well when he told Bloomberg, “Our American peers started earlier, but we’re catching up really fast.”
The Risks Investors Can’t Brush Off
Still, there are plenty of unknowns. Baidu hasn’t revealed per-ride revenue or detailed AV operating costs; everything remains bundled within broader business units. Until those numbers are separated, skeptics can argue that Apollo Go might just be an expensive growth experiment dressed up with glossy stats.
International expansion adds more complexity. Partnerships like the one with Switzerland’s PostBus—targeting full operations by early 2027—and pilot programs in Dubai and Abu Dhabi sound exciting but won’t move the profit needle soon. Between data rules, local safety tests, and partner economics, meaningful returns may not appear before 2028.
Export restrictions pose the biggest threat. U.S. controls on advanced semiconductors could push up Baidu’s sensor and training costs, cutting into the very cost advantage that sets it apart. And one major safety incident could quickly freeze city permits and derail expansion.
What’s Next
The near-term signals are clear. Baidu’s next earnings report will show whether it officially includes Apollo Go’s metrics or keeps them in the PR bucket. Any disclosure like “Wuhan achieved positive margins in Q4” would confirm profitability and silence doubters.
Keep an eye on fleet size too. If Baidu expands from 1,000 to 1,500 or even 2,000 cars by 2026, it’s doubling down on its lead. A foreign licensing or revenue-guaranteed partnership—something akin to Uber’s model—could also ease concerns that Apollo Go’s success is limited to China.
Here’s what’s undeniable: China’s dense cities and centralized regulation make it easier for robotaxis to thrive than in the U.S., and Baidu’s cost advantage gives it breathing room that Western rivals can’t match. Whether that advantage turns into real investor value depends on one thing—transparency.
But one fact is crystal clear. Baidu now operates at a scale big enough to demand it.
NOT INVESTMENT ADVICE