Tesla announces Robotaxi expansion to two new Florida cities while its actual fleet shrinks in every existing market

Tesla Is Adding Robotaxi Cities While Its Fleet Shrinks Everywhere It Already Runs

Tesla picked earnings day to announce that its Robotaxi service is expanding to Orlando and Tampa. The headline reads like growth. The fleet data says otherwise: in every city where Tesla currently operates driverless rides, the number of unsupervised vehicles has gone down.

The company said in a post on X that it was bringing Robotaxi service to the two Florida markets and included maps showing each city's service area. It did not say how many vehicles would deploy or whether riders could book a trip immediately 1. The post arrived hours before Tesla's Q2 earnings call, where CEO Elon Musk was expected to brief investors on Robotaxi progress.

The numbers behind the announcement are the part that matters. In May, Tesla had 29 unsupervised robotaxis in Austin, five in Dallas, and six in Houston, according to the Robotaxi Tracker, an independent platform that monitors autonomous vehicle deployments 2. As of this week, those counts stand at 17 in Austin, four in Dallas, and zero in Houston 1. Tesla has not explained the declines.

The gap between these numbers and what Musk has publicly promised is wide. He projected "500 or more" vehicles in the Austin area alone 1. Austin has 17 unsupervised vehicles. Across all three active markets combined, Tesla operates roughly 21 unsupervised robotaxis. For comparison, Waymo runs an estimated 3,500 fully driverless vehicles across more than 10 cities 1. Tesla's total unsupervised fleet amounts to less than one percent of that figure.

The rollout has been inconsistent in other ways. When entering new markets, Tesla usually starts with a human safety driver at the wheel. In some cases, the company has relocated that person to the passenger seat, giving them a kill switch if problems arise. Fleet counts in individual cities have fluctuated without explanation from the company 1.

This week also produced a new incident. Tesla disclosed that one of its unsupervised robotaxis struck a tree stump during a remote operation session 1. A vehicle operating without an in-car driver, steered by a remote operator, hitting a stationary object is the kind of event that compounds concerns about scaling unsupervised rides before the technology can run them reliably at volume.

For investors calibrating Tesla's autonomy timeline, the mismatch between announcement cadence and fleet trajectory is the figure that should weigh heaviest. Musk has projected that the Robotaxi service will turn profitable in 2027 1. That outcome depends on building ride density in markets where Tesla already operates. It does not depend on adding city names to a service map while the vehicles already deployed quietly disappear. A company can announce new markets faster than it can field working vehicles in the ones it already claims to serve. Tesla is currently doing the former while the latter moves in the wrong direction.

The pattern is consistent across every existing market. Each city has fewer unsupervised vehicles now than it did in May, and Houston has lost its fleet entirely. Orlando and Tampa will generate coverage when they open. Whether their fleets will still be running in six months, or whether they will follow Houston toward zero, is a question Tesla has not addressed.