Tesla launched Robotaxi service in Dallas and Houston on Saturday, April 18, via a post on the @robotaxi X account that read "Robotaxi now rolling out in Dallas & Houston" — no fleet-size disclosure, no pricing, no indication of whether the rides would be supervised or unsupervised. [1] Elon Musk reposted the announcement from his personal account with "Try Tesla Robotaxi in Dallas & Houston!" By Sunday afternoon, twenty-four hours after the service activated, Electrek's Robotaxi Tracker had crowdsourced a verdict that does not match the announcement. Dallas and Houston are running at 0 to 2 percent availability. A single vehicle has been reported by riders in each market. [2] Austin, the pilot city, has 46 vehicles in service and 19 to 20 percent availability — a baseline the new expansion is not meeting.

The paper's Sunday coverage of the Saturday launch as a narrative hedge three days before Q1 earnings named the structural question: whether expansion into two markets that Waymo has served with fully driverless vehicles since February 2026 would be distinguishable from a pre-print news cycle produced for bankers and a CEO rather than riders and regulators. The Electrek data answers that question cleanly. Tesla's Dallas geofence covers roughly 30 to 35 square miles centered on Highland Park and Park Cities. Houston's covers 12 to 15 square miles in the Willowbrook-Jersey Village area of the northwest side. [3] The Houston metropolitan area is larger than 10,000 square miles. The two new markets, measured by area served, are a rounding error against the cities they ostensibly cover.

Fred Lambert, the Electrek editor-in-chief who has covered Tesla's autonomy program for a decade, published the Sunday status check that framed the weekend rollout: "Tesla launched its 'Robotaxi' service in Dallas and Houston yesterday — expanding to two new cities for the first time since Austin. There's just one problem: there are virtually no cars available." [2] The Robotaxi Tracker's data shows availability reaching approximately 50 percent for narrow morning windows before returning to zero. Lambert's February status check on the Austin pilot after eight months of service found 42 vehicles, 19 percent availability, and a crash rate "9 times worse than human drivers" based on NHTSA Standing General Order incident data. [4] The Austin baseline is the context Dallas and Houston will be measured against; the new markets' one-car-each data point is not a decimal place off Austin, it is two orders of magnitude below.

Waymo provides the counterfactual. Alphabet's autonomous vehicle unit has operated fully driverless service in both Dallas and Houston since February 2026 through an Avis Budget Group partnership — no safety monitors, no chase cars, no remote supervision required. [3] Waymo is delivering 500,000 paid rides per week across ten U.S. cities, on its way to a target of one million. Tesla announced its Texas expansion into markets Waymo was already serving at scale, without disclosing fleet size, and delivered Sunday data of one vehicle per city. The comparison is the basis for Reuters's observation in its Saturday filing that "Alphabet's Waymo and Amazon's Zoox speeding up" is what the robotaxi business is doing while Tesla is trying to catch the narrative. [5]

Tesla reports first-quarter results after market close on Wednesday, April 22. Production in the quarter was 408,386 vehicles; deliveries were 358,023. The 50,363-unit inventory overhang is the largest in four quarters. [2] Consensus estimates from the Street have Tesla at approximately $0.39 EPS on $22.96 billion of revenue; some sell-side desks have Q1 operating margin back near 5 percent against Q1 2025's 2.1 percent. JPMorgan's most recent recorded target is in the $145-range with an Underweight rating. The stock is down approximately 20 percent year-to-date at a forward price-to-earnings multiple that sell-side calls 427. A 50,000-unit inventory build three days before the call is what the Saturday robotaxi rollout was visibly announced against.

The Saturday rollout's defense, articulated on Tesla-bull X by accounts like @WholeMarsBlog, is that expansion does not need to run at scale on day one to be real; Austin took eight months to reach 42 vehicles and 19 percent availability. If that is the ramp curve, Dallas and Houston are not underperforming expansion, they are executing expansion — just earlier in the curve. The problem with that defense, which Electrek's data undercuts, is that Austin's ramp curve is itself not what Musk promised. [4] Promises on record from Musk through 2025: 500 cars in Austin by end of year; half the U.S. population served; expansion to 8 to 10 metro areas by year-end. Actual delivered at end of 2025: 42 cars in Austin, 19 percent availability, two cities total.

What makes the pattern pre-earnings rather than operational is what Tesla did not do. The company did not disclose fleet size. It did not disclose whether rides would be supervised or unsupervised (its own X post included video of a Model Y with no one in the front seats, but a February 13 CPUC filing revealed Tesla still uses both in-car human drivers and domestic remote operators). [4] It did not disclose pricing. It did not explain why Dallas's geofence is shaped like an irregular trapezoid that includes much of downtown but cuts out Oak Cliff. The service still shuts down when it rains; Houston averages over 100 rainy days per year. The announcement exists. The service barely does.

Wednesday's call will be asked for three specific things by equity analysts: the inventory overhang plan, the margin trajectory against the $145-target bear case, and the robotaxi fleet-size disclosure the Saturday announcement did not produce. The tell to watch is which of those three Musk discloses numerically versus which he reframes as narrative. If the robotaxi number on Wednesday is a fleet-size the post-announcement ramp can plausibly deliver in a quarter, the Saturday announcement was operational. If the number is "we'll expand to Miami, Phoenix, Tampa in the first half" — the January promise already missing a month of Q2 — the Saturday announcement was the narrative, and Electrek's one vehicle per city is the news.

The second-order read the paper's bank-war-economy thread has been tracking is whether narrative hedges work on a peace-priced market that has just shown it will punish growth-ceiling admissions. Netflix demonstrated Friday that the market is not awarding growth credit for stories that cannot defend their earnings trajectory. SpaceX's analyst day on Tuesday tests whether the same market will underwrite the biggest IPO in history at 109 times EBITDA. Tesla's Wednesday call tests whether the Saturday robotaxi story can substitute for the 50K inventory reality. Electrek's Robotaxi Tracker has already provided Monday's answer in miniature: the service exists on X and does not yet exist on the street.

-- THEO KAPLAN, San Francisco