Noah Romsdal Hallundbæk Sørensen
Tesla’s robotaxi fleet has fallen to just 20 active vehicles, and the distance between the company and rival Waymo grows wider with every passing week.
Just one month after signs of progress, Tesla’s self-driving fleet is now smaller than ever.
The number of active, unsupervised Tesla robotaxis has dropped to just 20 vehicles, down from 25 at the end of April.
Looking at Tesla’s entire ride-hailing operation, the picture is even more discouraging, with only 34 active vehicles across all locations.
According to Electrek
The numbers are moving in the wrong direction
New data from Robotaxi Tracker shows that the unsupervised fleet now consists of 14 active vehicles in Austin, 3 in Dallas and 3 in Houston.
That is a drop from the 25 vehicles Electrek reported in April, and the brief growth seen at the time now appears to have been a temporary peak rather than a genuine upward trend. Austin, which has served as the flagship city for the programme, has lost five active vehicles in under a month.
Dallas and Houston have not grown at all since the April launch and remain fixed at the same three vehicles each. The fleet Tesla presented as evidence of progress has therefore already shrunk before the programme properly got off the ground.
Bay Area collapse drags the numbers down
The Bay Area fleet, which consisted of supervised Full Self-Driving vehicles, has fallen from 107 active units in April to just 9 today.
That is a dramatic collapse for the part of the operation that previously made up the vast majority of Tesla’s ride-hailing capacity.
These cars never operated as true robotaxis, as they were driven with safety personnel behind the wheel under California’s transportation permit.
The total active fleet peaked around December 2025 and January 2026 and has been declining ever since. The brief recovery in unsupervised vehicles reported in April looks, in hindsight, like an exception rather than a trend.
Safety is acting as a brake on growth
Tesla has offered no official explanation for the decline, but the likely answer is the same one Electrek has reported for months: safety is the bottleneck.
Elon Musk himself has acknowledged to investors that safety validation is the limiting factor, and Tesla’s crash rate with unsupervised vehicles has been reported to be approximately four times higher than that of human drivers.
More cars mean more miles, and more miles mean more incidents.
Musk has pointed to FSD v15, an upcoming software rewrite, as the solution, but that pushes any meaningful scaling to late 2026 or early 2027. Until then, the strategy appears to be keeping the fleet small rather than risking further incidents.
Waymo continues to push forward
While Tesla contracts, Waymo is expanding. Alphabet’s self-driving company now operates around 3,000 robotaxis across several US cities and completes hundreds of thousands of paid trips every week.
Waymo is also constructing a new production facility in Mesa, Arizona, and is preparing launches in Atlanta, Miami and Washington D.C. later this year.
The contrast between Tesla’s 20 active unsupervised vehicles and Waymo’s thousands makes the scale of the gap strikingly clear. Waymo is not without its own challenges, however, and recently had to suspend its service on motorways and in select markets due to difficulties detecting flooded road surfaces.
