Waymo is pulling away from Tesla in the robotaxi race, with the Alphabet-backed company now operating approximately 3,000 robotaxis across multiple U.S. cities while Tesla's unsupervised fleet has shrunk to just 20 active vehicles. The contrast illustrates how differently the two companies are approaching autonomous vehicle deployment, with Waymo scaling aggressively while Tesla appears to be prioritizing safety constraints over growth.

Why Is Tesla's Robotaxi Fleet Getting Smaller?

Tesla's unsupervised "Robotaxi" fleet has declined dramatically from 25 cumulative vehicles reported in late April to just 20 active vehicles in the most recent data. The collapse is even more severe when looking at the entire ride-hailing operation, which has dropped from 165 total active vehicles in April to just 34 vehicles across all locations. Austin, which was supposed to be the flagship deployment city, lost five vehicles, falling from 19 to 14 active in the last seven days. Dallas and Houston remain stuck at just 3 vehicles each, unchanged since their April launch.

The likely culprit behind this contraction is safety. Tesla's crash rate with unsupervised vehicles has been reported at roughly four times worse than human drivers. Elon Musk himself told investors that safety validation is the limiting factor preventing fleet expansion. More vehicles mean more miles, and more miles mean more crashes, so Tesla appears to have made the rational decision to pull vehicles back rather than add them.

Musk has indicated that Tesla is waiting for improvements from a rewrite reaching consumers in Full Self-Driving (FSD) v15 before scaling the robotaxi fleet aggressively, pushing the timeline to late 2026 or early 2027. In the meantime, the fleet is getting smaller, not bigger.

How Is Waymo Taking a Different Approach?

While Tesla's fleet shrinks, Waymo is expanding rapidly. The company now operates approximately 3,000 robotaxis across multiple U.S. cities, completing hundreds of thousands of paid trips per week. Waymo is building a new manufacturing facility in Mesa, Arizona with Magna to produce over 2,000 additional vehicles, and it is preparing to launch in Atlanta, Miami, and Washington, D.C. later this year.

Waymo's vehicles operate in more cities, with broader geofences and higher utilization rates compared to Tesla's limited deployments. The gap between Tesla's 20 active unsupervised vehicles and Waymo's thousands illustrates not just a difference in scale, but a widening gulf in execution capability.

However, Waymo is not without its own challenges. The company recently had to pause service on highways and in several markets due to issues with vehicles not always detecting flooded streets. Service was paused in Atlanta, San Antonio, Dallas, and Houston after vehicles entered flooded intersections despite previous recall precautions. Waymo also paused freeway service in response to a high-speed incident and apparent police chase in a construction zone.

What Are the Key Differences in Deployment Strategy?

• Scale and Utilization: Waymo operates 3,000 robotaxis completing hundreds of thousands of paid trips weekly, while Tesla operates just 20 unsupervised vehicles with minimal utilization across three cities.
• Geographic Expansion: Waymo is expanding to Atlanta, Miami, and Washington, D.C. this year and building manufacturing capacity for 2,000 additional vehicles, while Tesla's fleet is contracting in its existing markets.
• Safety Approach: Tesla is prioritizing safety validation over growth and pulling vehicles off the road, while Waymo is scaling despite encountering safety issues that require recalls and service pauses.
• Manufacturing: Waymo is partnering with Magna to produce vehicles at scale, while Tesla relies on its own production capacity, which has not translated into fleet growth.

How Are Technology Partnerships Shaping the Autonomous Vehicle Industry?

The divergence between Tesla and Waymo reflects a broader industry trend toward strategic partnerships in autonomous driving. According to recent analysis, automakers and ride-hailing companies are increasingly recognizing that close collaboration with technology leaders, especially chipmakers, is essential for scaling autonomous systems.

These partnerships go beyond traditional supplier relationships. For example, NVIDIA and Mercedes-Benz are collaborating on joint research and development in vehicle intelligence, simulation environments, and industrialized deployment processes. Similarly, NVIDIA and ride-hailing provider Bolt are building an open-source autonomous vehicle platform that combines computing, simulation, and software with real-world fleet data.

"The AI-driven transformation is accelerated when OEMs co-develop with chipmakers across the full lifecycle," stated Stefan Riederle, Senior Partner at Roland Berger.

Stefan Riederle, Senior Partner, Roland Berger

Automakers require three key capabilities to achieve scale in autonomous driving: advanced perception powered by artificial intelligence (AI) to interpret complex environments and recognize hazards; real-time decision-making enabled by high-performance computing platforms that process vast amounts of data in milliseconds; and adaptive learning through AI models that continuously improve through data-driven training.

What Does This Mean for the Future of Robotaxis?

The contrast between Tesla and Waymo raises fundamental questions about how to balance safety with deployment speed. Tesla's approach of shrinking its fleet to address safety concerns is defensible from a risk management perspective, but it undermines the business case that has supported the company's valuation. Waymo's willingness to scale despite encountering safety issues, followed by recalls and service pauses, suggests a different calculus: deploy at scale, identify problems in the real world, and fix them through software updates.

However, experts are questioning whether either approach adequately addresses the full spectrum of safety concerns. A single metric like crashes per mile does not capture all safety issues, including vehicles blowing past stopped school buses, driving into flood waters, blocking fire trucks, or navigating construction areas.

"A net statistical number isn't the whole answer," noted Phil Koopman, an autonomous vehicle safety expert.

Phil Koopman, Autonomous Vehicle Safety Expert

As robotaxi companies mature, the industry will need to move beyond simple safety metrics and develop comprehensive approaches that satisfy multiple stakeholders, including regulators, safety advocates, and the public. Tesla's bet on FSD v15 solving its problems remains exactly that, a bet. Until the software is dramatically better, the fleet will stay small, and the gap with Waymo will only grow.