Tesla Robotaxi Now Covers 7 Cities: What the Florida Blitz Tells Us About the Race to Scale

Eighteen days. That's how long it took Tesla to go from its Miami Robotaxi launch to operating fully driverless rides in Orlando and Tampa as well. As of July 21, Tesla's autonomous ride-hailing service is live in seven US cities across three states — and the pace of expansion is accelerating in a way that would have seemed implausible just six months ago.

But speed isn't the only story. The Florida push is also the most technically demanding thing Tesla has attempted with its camera-only self-driving system — and regulators, rivals and riders are all watching closely to see if the technology holds up where the weather does not cooperate.

What Just Happened

Tesla launched its Robotaxi service in Miami on July 3, 2026 — the first driverless ride-hailing service outside Texas and California. Then, on July 21, it flipped the switch on Orlando and Tampa simultaneously, extending its Florida footprint to three metros in under three weeks.

What makes this expansion notable isn't just the speed. It's the operational decision Tesla made in Miami that it has now carried forward: going fully driverless from day one, with no human safety monitor in the front seat, in a brand-new city where the system had never operated commercially before. That's a departure from how every prior Tesla Robotaxi market launched — Austin, Dallas, Houston and the Bay Area all began with supervised rides before transitioning to unsupervised operation. Miami broke that pattern. Orlando and Tampa followed the same model immediately.

Tesla's head of AI, Ashok Elluswamy, confirmed on social media that Miami rides are running fully unsupervised. Early riders posted footage of Model Y vehicles navigating Florida streets with empty front seats — no driver, no monitor, no one to take over.

Tesla's Robotaxi network now covers Austin, Dallas, Houston, the San Francisco Bay Area (still with a safety monitor), Miami, Orlando and Tampa. For a wider view of how Tesla stacks up against its main rival, see our Tesla Robotaxi vs. Waymo head-to-head comparison.

The Florida Challenge: Rain, Glare, and a Camera-Only Bet

Tesla's camera-only approach — no lidar, no radar — has always had a specific weakness that engineers and critics have flagged repeatedly: degraded visibility conditions. Lidar doesn't care about rain. Cameras do. We unpacked that trade-off in Lidar vs. Cameras: The Sensor Shootout Defining Autonomy.

Florida is not the place you'd choose to test a camera-only system if you had any choice about it. The state averages more than 130 thunderstorm days per year in summer. Tropical downpours can reduce visibility to near zero in seconds. Intense sun glare off wet roads creates exactly the kind of optical interference that camera systems struggle with most.

NHTSA made this concern official in March 2026, when an engineering analysis found that Tesla's FSD system "fails to detect and/or warn the driver appropriately under degraded visibility conditions such as glare and airborne obscurants." That analysis covered FSD Supervised on consumer vehicles — but the same underlying camera architecture powers the Robotaxi service. Our honest review of Tesla FSD in 2026 goes into how that plays out for everyday drivers.

Tesla's counter-argument is its training data. The company has accumulated over 10 billion FSD miles globally, including hundreds of millions in varied weather conditions. Its end-to-end neural network learns from edge cases, and every commercial mile in Miami — every sudden downpour, every blinding afternoon glare — generates training data that feeds back into the model. The architecture behind that pipeline is explained in Inside Tesla FSD v13: The End-to-End Neural Net Explained.

The next four months of Florida summer will be the most direct test yet of whether that argument holds in commercial operation at scale.

The Map vs. Fleet Problem

One consistent pattern across Tesla's Robotaxi expansion is the gap between how big the service area looks on a map and how many cars are actually in it.

Austin expanded to cover the entire metro — 245 square miles — on June 3, 2026. The fleet serving that area at launch: approximately 20 vehicles. Miami's service zone covers roughly 10 to 14 square miles of western Miami-Dade. Tesla has not disclosed the fleet size there.

On Tesla's Q1 2026 earnings call, Elon Musk explained the logic: safety validation is the binding constraint, not geography. The company is deliberately holding back aggressive fleet deployment until it is confident in the system's performance in each environment. A wide map with a small fleet is a feature, not a bug — it lets Tesla gather data across a broad area while limiting exposure in any single location.

The practical effect for riders is that availability is patchy. Unlike Waymo, where a fleet of roughly 2,500 vehicles across 11 cities means consistent availability, Tesla's smaller fleet means waits can be long and coverage uneven within the service area. For context on Waymo's scale, see Waymo Expands Robotaxi Service to Five New Cities in 2026 — and on the risk side, Waymo Recalls 3,871 Robotaxis After Cars Drove Into Active Construction Zones.

Tesla has said large-scale fleet expansion will follow the release of FSD version 15 — a major model rewrite growing from roughly 1 billion to 10 billion parameters — which Musk has pointed to as arriving in late 2026 or early 2027.

The Cybercab Is Still Waiting

Florida's three new cities are running on modified Model Y vehicles — not the Cybercab, Tesla's purpose-built autonomous two-seater with no steering wheel and no pedals. The Cybercab began production earlier this year and has been spotted in limited Austin testing, but commercial deployment remains on hold pending regulatory clearance.

The reason: under current federal rules, vehicles without steering wheels require an NHTSA exemption before they can operate commercially, capped at 2,500 units per manufacturer annually. NHTSA's administrator signalled earlier this month that steering wheel requirements may eventually be dropped altogether — we covered that policy shift in The Steering Wheel Is Next: NHTSA Signals the End of a Century-Old Requirement — but that rule change is 12 to 24 months away at minimum. Until then, Tesla's robotaxi fleet stays on Model Ys.

The Cybercab's economics are the real prize. Tesla has said it targets a manufacturing cost of around $30,000 per unit — dramatically below the $100,000-plus cost of purpose-built lidar-equipped competitors like Zoox. If and when the Cybercab gets commercial clearance, Tesla's unit economics improve sharply. Florida is meanwhile building the operational playbook the Cybercab fleet will eventually use.

How This Compares to Waymo

The contrast with Waymo remains stark — and worth being precise about.

Waymo operates roughly 2,500 fully driverless vehicles across 11 US cities, completing more than 500,000 paid rides per week. Its safety data, covering over 220 million autonomous miles, shows 94% fewer crashes causing serious or fatal injuries compared to human drivers in the same areas.

Tesla has not published equivalent safety data for its Robotaxi service. It has not disclosed fleet size, total miles driven in unsupervised mode, or incident rates specific to the robotaxi program. NHTSA has an open investigation into Tesla's FSD system, and separate from that, prosecutors in Texas recently charged a driver with manslaughter following a fatal crash in which the vehicle's telemetry showed a manual override — underscoring how carefully regulators are watching Tesla's autonomous operations.

What Tesla does have that Waymo doesn't is speed. Waymo took years to reach 11 cities. Tesla added three cities in 18 days. The question the industry is debating is whether the pace reflects genuine technical confidence or an aggressive market-capture strategy that is outrunning the evidence base.

What's Next

Tesla has flagged Phoenix and Las Vegas as next in line, with Nevada permits already filed covering up to 5,000 vehicles in year one. Elon Musk has said he targets Robotaxi service across a dozen US states by year-end — which would mean roughly five more state launches in five months.

Whether that timeline holds depends on three things: FSD v15's readiness, NHTSA's ongoing scrutiny of the existing fleet's safety record, and how the Florida summer goes. If Miami, Orlando and Tampa post clean safety records through hurricane season, the case for rapid further expansion becomes considerably easier to make. If there are incidents — particularly weather-related ones — the regulatory and public pressure to slow down will intensify.

Europe remains a separate question. Tesla FSD is approved in the Netherlands and under active evaluation in Belgium, but the Robotaxi service specifically requires a different and higher level of regulatory clearance that no European country has yet issued. The EU's autonomous vehicle framework, coming into effect in January 2027, is the most likely pathway — see our EU AV Act explainer and Europe's self-driving legal landscape guide for the full picture.

For now, Florida is the story. And Florida's weather will have more to say about Tesla's robotaxi future than any earnings call or regulatory filing.