BIDU, PONY and WRD: In the Driving Seat of China’s Robotaxi Boom

Nightmare scenario: on April 1, at least 100 robotaxis belonging to Baidu’s [BIDU] Apollo Go simply stopped working on the streets of Wuhan, leaving passengers trapped inside. 

Apollo Go is China’s biggest robotaxi operator and, as such, the incident caused considerable consternation.

On April 14, the Chinese Ministry of Industry and Information Technology duly announced that it had instructed local authorities to carry out self-inspections and tighten safety supervision of road testing following the incident. 

Nevertheless, the robotaxi genie is out of the bottle in China. Even a glitch of this magnitude won’t halt its progress.

According to a 2025 report from Goldman Sachs, China’s robotaxi market is entering a high-growth phase, with projections estimating a potential 700x growth from 2025 levels to a market size of $47bn by 2035, driven by falling hardware costs and government support.

Three firms are at the forefront of the space.

Baidu is the clear leader, thanks to Apollo Go, which is already delivering large-scale paid rides and expanding overseas. Pure-play names Pony.ai [PONY] and WeRide [WRD], meanwhile, are scaling fleets across China and the Middle East. Collectively, these firms are pushing Level 4 autonomy towards commercial viability, with China now hosting more active robotaxi deployments than the US.

Let’s look at each of these three big names in turn, then see how their fundamentals compare. 

Baidu: The international expansion play

Baidu has evolved its Apollo Go platform from pilot-stage testing into one of the world’s largest commercial robotaxi networks. The fleet is active across major Chinese cities including Wuhan, Beijing and Shenzhen, with deployments now spanning more than 20 cities. It has also begun early international expansion, including pilots in markets such as Dubai and South Korea. Recent data indicates tens of millions of fully driverless rides have been completed, with weekly volumes exceeding 300,000 trips.

The operating model is increasingly designed for replication, featuring standardised RT6 vehicles, a centralised autonomous driving stack, and a city-by-city rollout framework that reduces marginal deployment cost. Baidu is transitioning from safety-driver assisted operations towards fully driverless commercial zones, particularly in dense urban districts where mapping and regulatory approval are most mature.

Strategically, Baidu is the benchmark scaled player in China’s robotaxi market. The longer-term thesis increasingly mirrors a platform export model, with partnerships with mobility aggregators such as Uber [UBER] and Lyft [LYFT] acting as potential international distribution channels. Key constraints remain regulatory fragmentation and intermittent system-level safety incidents, but improving utilisation is steadily strengthening unit economics.

Pony.ai: The asset-light play

Pony.ai represents the most direct US-listed pure-play exposure to China’s robotaxi scale-up story. The company operates seventh-generation autonomous vehicles across Beijing, Shanghai, Guangzhou and Shenzhen, with service areas expanding into dense urban cores and transport nodes such as airports and railway corridors.

Unlike Baidu’s vertically integrated ecosystem approach, Pony.ai is more explicitly asset-light, partnering with original equipment manufacturers (OEMs) and fleet operators to scale deployment without fully owning vehicle infrastructure. It is targeting more than 3,000 robotaxis in 2026, up from fewer than 1,000 at the end of 2025: a sharp inflection in commercial ambition.

Internationally, Pony.ai is following a similar trajectory to Baidu but at earlier scale, extending into select Middle Eastern and European markets. Early unit economics improvements in specific Chinese cities suggest that higher ride density and utilisation are beginning to offset deployment costs. Strategic partnerships with OEMs such as Toyota [TM] and BAIC [BMCLF] reinforce its positioning as a scalable autonomy stack rather than a single-market operator. The key execution variable is whether it can replicate city-level profitability consistently across jurisdictions.

WeRide: The diversified platform play

WeRide is in a slightly different place to its peers, positioning itself as a multi-vertical autonomous driving platform rather than a pure robotaxi operator. Its robotaxi operations are live in cities such as Guangzhou and Beijing, focused on highly mapped zones where full autonomy has regulatory approval.

Relative to Baidu and Pony.ai, WeRide’s differentiation lies in breadth: its autonomous stack spans robotaxis, robobuses and logistics vehicles, which diversifies revenue pathways but dilutes pure-play exposure. It remains smaller at robotaxi scale, but is increasingly using international markets to accelerate credibility and regulatory optionality.

A key strategic anchor is Abu Dhabi, where WeRide operates one of the most advanced commercial robotaxi deployments outside the US and China. This overseas footprint is central to its thesis: proving that its system can be replicated under different regulatory regimes, reducing dependence on Chinese rollout cycles.

Commercially, WeRide is pushing an asset-light expansion model through OEM partnerships, including with Geely [GELYF], while integrating ride-hailing distribution via platforms such as Uber to improve demand capture. The investment case ultimately hinges on whether its diversified autonomy strategy can still deliver robotaxi-scale economics comparable to more focused peers.

BIDU vs PONY vs WRD

This is how the three stocks currently compare.

Source: Yahoo Finance

Conclusion: Investing in Chinese robotaxis

China’s robotaxi sector is transitioning from experimental deployment to early-stage commercial infrastructure, with Baidu setting the scale benchmark. The investment case across the stocks analysed here is less about technological feasibility and more about execution on three variables: utilisation, regulatory throughput and unit economics.

On the positive side, China offers structural advantages that are difficult to replicate elsewhere, in the form of dense urban environments with large ride-hailing demand pools – not to mention strong state coordination. This is already visible in Baidu’s rising trip volumes and early signs of per-vehicle profitability emerging at the margin for smaller operators. International expansion, particularly into the Middle East, adds a second growth vector and helps de-risk domestic policy cycles.

However, the path to durable profitability remains uneven. Regulatory fragmentation across cities, intermittent safety interventions and the capital intensity of scaling fleets continue to weigh on returns. Even with asset-light models and OEM partnerships, the economics depend heavily on sustained utilisation gains that are still early in development.

Overall, the sector resembles a land-grab phase where scale leaders may eventually consolidate economics, but execution risks remain high and a clear gap between winners and laggards has yet to emerge.

CMC Aureon’s proprietary theme relevance system maps the world’s biggest investing megatrends. For in-depth analyses of stocks with high growth potential, subscribe to CMC Aureon Foresight.