Tesla Robotaxi Goes Unsupervised in Dallas and Houston With Five Cities Left to Launch

Photo by Trac Vu on Unsplash

Tesla launched fully driverless Robotaxi service in Dallas and Houston on April 18, 2026, and for the first time skipped the supervised-rides phase that defined its Austin pilot. The Tesla robotaxi unsupervised expansion was confirmed by Tesla's official Robotaxi account and corroborated by user-shared video of empty front seats, as documented by Not a Tesla App, which also noted Tesla's break from its Austin playbook where supervised rides preceded unsupervised ones.

The rollout lands four days before Tesla's Q1 2026 earnings call, with five cities still owed under the company's H1 expansion plan and a confirmed fleet of just 25 driverless vehicles across all three Texas markets.

A 25-Vehicle Fleet Behind a Trillion-Dollar Narrative

The scale of the operation is the part the press release does not advertise. As of late April and early May 2026, Tesla's total confirmed unsupervised Robotaxi fleet across three Texas cities stands at just 25 vehicles, with 19 in Austin, 3 in Dallas, and 3 in Houston, according to Electric Vehicles.

That is the same publication's count. A separate tally from Basenor puts the active unsupervised fleet at roughly 20 vehicles in May, with 14 in Austin rather than 19, suggesting Tesla has been pulling cars back rather than scaling up. Either number is a rounding error against Waymo.

Tech Insider puts Waymo at roughly 3,000 robotaxis across 10 U.S. cities serving more than 500,000 weekly rides, versus a broader Tesla footprint of about 573 vehicles across 3 cities once supervised cars are counted. That is a vehicle-count gap of more than 5:1, and the unsupervised gap is closer to 100:1.

The Houston geofence covers roughly 12 to 15 square miles around Jersey Village and Willowbrook. Dallas launched with a significantly larger 30 to 35 square mile footprint covering the urban core and Park Cities.

Why Skipping the Safety Monitor Matters

Tesla's Austin pilot used a human in the front passenger seat for months before that person disappeared. Dallas and Houston got neither. The Not a Tesla App report calls the decision a surprising move, and it is, because the regulatory and reputational cost of a single high-profile incident in a new market with no in-car monitor is asymmetric.

The NHTSA-reported record is not blank. Two confirmed safety incidents have surfaced: in July 2025 in Austin, a teleoperator took control after the ADS stalled, mounted a curb, and struck a metal fence; and in January 2026, a teleoperator took over and the vehicle collided with a construction barricade at approximately 9 mph, scraping the front-left fender, per Teslarati.

Both incidents involved remote intervention. Tesla has confirmed to lawmakers that teleoperators are authorized to pilot vehicles remotely but only at speeds below 10 mph, and only for repositioning in awkward situations, not for general navigation. The implication is the part nobody is selling: a driverless Tesla in Dallas or Houston is not fully autonomous in the engineering sense. It is autonomous up to the moment a remote operator in a chair somewhere has to nudge it out of a parking lot or off a curb. That is closer to how Waymo runs its fleet than Tesla's marketing suggests, and it is the single most important caveat to the word "unsupervised."

Tesla Robotaxi vs Waymo in the Same ZIP Codes

The Dallas and Houston launches create the first apples-to-apples Tesla robotaxi vs Waymo comparison on identical streets. Waymo began operating in Dallas, Houston, San Antonio, and Orlando in February 2026, meaning Tesla entered four-month-old Waymo territory rather than virgin markets.

The sensor philosophies diverge at the bill of materials. Waymo uses LiDAR, cameras, and radar, while Tesla relies solely on a camera-based vision system. That is the core architectural difference shaping both unit cost and the safety debate, and it is why Tesla can plausibly target sub-$30,000 hardware while Waymo's Jaguar I-Pace sensor stack reportedly costs six figures per vehicle.

Price is the other front. During the Austin pilot, Jefferies analysts found Tesla Robotaxi fares were discounted approximately 60 percent compared with UberX prices, a level of subsidy that only works if Cybercab unit economics eventually arrive. Tesla has not officially disclosed pricing for Dallas or Houston.

Tesla FSD Unsupervised Launch Cities 2026: The Five Still Outstanding

The H1 expansion plan named seven cities. Three are live. Tesla still has five cities remaining on its H1 2026 expansion plan, Phoenix, Miami, Orlando, Tampa, and Las Vegas, with roughly two months left to hit the deadline, per Electric Vehicles.

The hiring data lines up with that map. Tech Insider reports Tesla has been recruiting fleet support specialists across nine cities: Austin, Dallas, Houston, Las Vegas, Tempe, Palo Alto, Tampa, Doral, and Orlando. Doral covers the Miami metro, Tempe covers Phoenix, and Palo Alto is the regulatory holdout.

The California problem is structural. The Bay Area fleet exceeds 500 vehicles but still requires a safety driver due to state regulations, limiting unsupervised commercial operations there. That single line explains why Tesla's biggest vehicle pool generates the least revenue per car: it cannot legally run driverless rides for paying passengers without a CPUC permit that Tesla has not obtained.

Musk's Quietest Admission on the Q1 Call

The most consequential sentence of the quarter was not the launch announcement. It was the guidance walk-back. On the Q1 2026 earnings call, Musk said he does not think unsupervised FSD or Robotaxi revenue will be super material this year, but does think it will be material in a significant way next year, according to Electric Vehicles.

For a stock that surged 12 percent over four trading sessions to close at $400.62 on April 17, 2026 ahead of the Dallas and Houston launch, with Wedbush analyst Dan Ives calling it a potential $3 trillion AI chapter for Tesla, "not material this year" is the line that matters. Musk also said unsupervised FSD for customer-owned vehicles would arrive probably in the fourth quarter of 2026, framed explicitly as a guess.

The Q1 shareholder letter itself flagged that paid Robotaxi miles nearly doubled sequentially during Q1 2026, which sounds large until the 25-vehicle denominator is factored in. Doubling a small number is still a small number.

Tesla Cybercab Release Date and the Production Reality

The Tesla Cybercab release date is no longer the bottleneck the hardware is. The Cybercab is a purpose-built two-seat autonomous vehicle with no steering wheel or pedals, unveiled October 2024, with volume production scheduled to begin April 2026 and a stated target price under $30,000, per Teslarati.

Production did start. On the Q1 2026 earnings call, Musk confirmed that Cybercab production had begun at Giga Texas, but tempered it: he warned investors to expect initial output to be "very slow" before any exponential ramp. Lars Moravy, Tesla's VP of Vehicle Engineering, told Electrek the Cybercab is self-certified to all existing FMVSS standards, meaning it sidesteps NHTSA's 2,500-unit annual exemption cap that constrains Waymo and Cruise.

That regulatory carve-out is the most underappreciated piece of the Cybercab story. If self-certification holds, Tesla's ramp ceiling is set by its factory, not by Washington.

What the Long-Tail Strategy Actually Is

Tesla VP of AI Software Ashok Elluswamy framed the company's safety approach in a way most coverage skipped past. Elluswamy stated the company is simultaneously solving the long tail of safety by monitoring the metrics across the entire Tesla customer vehicle fleet while scaling up its QA fleet nationwide.

Translated, that is the difference from Waymo. Waymo logs miles with its own cars in its own markets. Tesla logs miles with every Model 3, Model Y, Model S, and Model X running supervised FSD on every road in the country, then folds that telemetry back into the driverless stack. Whether that data advantage outweighs Waymo's LiDAR ground truth is the trillion-dollar question, and it is the one Morgan Stanley is implicitly betting on with its December 2025 forecast that Tesla will operate 1,000 Robotaxi vehicles by the end of 2026.

From 25 to 1,000 in seven months requires 40x growth. Anyone modeling Tesla's autonomy story should anchor on that gap, not the headline.

Evening Hours and the Slow Loosening of Constraints

The operational envelope is widening in increments rather than leaps. As of May 4, 2026, Austin's Robotaxi fleet began operating unsupervised during evening hours for the first time, having previously been restricted to daylight and mid-afternoon, while Dallas and Houston had supported evening unsupervised runs since their April launch.

The newer cities got fewer guardrails than the original. That is either confidence or pressure, and the answer depends on whether the next five launches arrive before June 30. For context on how other automakers are pacing their own hands-free rollouts, see our coverage of Stellantis and Wayve's 2028 North American timeline and how today's self-driving systems compare across brands.

The Date to Watch

June 30, 2026. That is the deadline Tesla set itself for the H1 expansion. Phoenix, Miami, Orlando, Tampa, and Las Vegas have to be live and driverless by then, or the milestone slips into the same H2 window Musk has already softened with his "probably Q4" guidance on consumer FSD. If even one of the five launches with a safety monitor in the seat, the Austin playbook is back and the Dallas-Houston model becomes the exception, not the new normal. More auto coverage lives in our Electric Vehicles and Connected Cars & ADAS sections.