Robotaxis in 2026: Where Self-Driving Works Now

Short answer: as of August 2026, Waymo offers paid driverless rides — no human behind the wheel — in nine US metro areas, WeRide runs a fully driverless fare service in Dubai, Baidu's Apollo Go covers 27+ Chinese cities, and Tesla's Robotaxi operates in about seven US metros but still keeps a safety monitor in the front seat on most rides.

Which cities actually have driverless robotaxis in 2026?

Waymo is the widest driverless operator in the US: Phoenix, the SF Bay Area, Los Angeles, Austin, Atlanta, and Miami were already live before 2026, and Waymo added Dallas, Houston, San Antonio, and Orlando in February 2026. London, Tokyo, Denver, and Washington DC are announced but not yet operating as of August 2026, according to CNBC's reporting on the expansion.

Outside the US, the picture is different. Baidu's Apollo Go runs in more than 27 Chinese cities, with full-city coverage in Wuhan, and has pushed into Dubai (via a partnership with Uber), Abu Dhabi, and Seoul, plus a 2026 testing permit in Switzerland. WeRide went further in one specific market: on March 31, 2026, it launched a fully driverless, fare-charging robotaxi service in Dubai through Uber, with no onboard operator at all. WeRide also runs in Abu Dhabi and Riyadh. Amazon's Zoox treats Las Vegas as its primary live market and got an NHTSA exemption to charge fares there, reported July 31, 2026.

Tesla's Robotaxi is the outlier on the "driverless" question. It launched in Austin in mid-2025 with an employee riding in the front passenger seat as a safety monitor, and by the first half of 2026 it had expanded to roughly seven metro areas — but on many of those trips, a human monitor is still present, so it isn't driverless in the same sense as Waymo or WeRide's Dubai service.

Which robotaxi operator is cheapest?

Tesla's Robotaxi is the cheapest by a wide margin: it averages about $8.17 per ride in the SF Bay Area, roughly half of Uber, Lyft, or Waymo, according to Sherwood News. The tradeoff is a much smaller fleet and longer wait times, since Tesla is still scaling up its vehicle count relative to established ride-hail networks.

Waymo has narrowed its price gap with human-driven ride-hail. TechCrunch reported in January 2026 that Waymo's average fare is about $19.69 versus roughly $17.47 for Uber and $15.47 for Lyft on comparable trips — a premium of about 13%, down from 30–40% in 2025. That's a meaningful shift: a year earlier, choosing Waymo over Uber routinely cost a third more; now it's closer to a rounding error for most riders who value not tipping or making small talk.

If you've ever waited twenty minutes for a cheap Tesla Robotaxi only to watch three empty Waymos glide past, you already know fleet size matters more than the sticker price. A $8 ride is not a deal if it never actually arrives.

How do Waymo, Tesla, and Zoox differ technically?

Waymo and Zoox both build their self-driving stacks around lidar — laser-based sensors that build a 3D map of surroundings in addition to cameras and radar. Tesla relies on a vision-only approach: cameras alone, with no lidar or radar, on the theory that if humans can drive with eyes only, cameras should suffice.

That split is now a public argument between the companies, not just an engineering footnote. In an August 4, 2026 piece, Waymo's co-CEO publicly criticized Tesla's camera-only approach as insufficiently redundant for safety-critical driving, per Electrek. Tesla has countered for years that lidar is an expensive crutch that cameras will eventually make unnecessary. Neither side has produced data that settles the argument publicly, and independent verification of comparative crash rates between the two approaches remains thin — treat any specific multiplier you see cited (like "Tesla crashes at three times the human rate") as unverified rather than established fact.

How are robotaxi safety incidents actually reported?

Robotaxi safety data comes primarily from self-reported disclosures to NHTSA plus follow-up regulatory probes, not from an independent, standardized incident database — so treat headline crash counts as a floor, not a full picture. Tesla disclosed 17 crash narratives tied to its automated driving system covering July 2025 through March 2026, per Electrek's review of unredacted filings.

Of those 17 incidents: 13 were property-damage-only, 2 involved no injury, 1 involved a minor injury, and 1 involved a minor injury that required hospitalization. Separately, in July 2026 a Tesla robotaxi under remote-operator control hit a tree stump at roughly 2 mph in Houston — a low-speed incident, but one that raised questions about how much human intervention remote operators actually provide during "driverless" trips.

Waymo faced its own scrutiny in 2026. On January 23, 2026, a Waymo vehicle struck a 9-year-old child near a school in Santa Monica, with minor injuries reported. NHTSA opened a probe, and its March 2026 follow-up request covered 12 areas, including how Waymo's vehicles handle school-zone protocols specifically. No operator has an unblemished 2026 safety record — the difference is in fleet size, incident type, and how much scrutiny regulators have applied so far.

What's blocking robotaxi expansion in 2026?

State and local regulation is now the main brake on expansion, more than the technology itself. New York Governor Kathy Hochul withdrew a robotaxi-legalization proposal, reported by TechCrunch in February 2026, keeping the state closed to the entire category regardless of operator.

New Jersey's bill S1677 would require both lidar and radar sensors on any robotaxi, which would effectively ban Tesla's camera-only design from operating in the state if it passes, according to Electrek. Washington DC's city council has separately discussed capping robotaxi fleet sizes and adding a 15-cent-per-mile tax on robotaxi trips specifically — a policy that would raise the cost of exactly the service model these companies are trying to scale. At the federal level, NHTSA has a proposal on the table, not yet finalized, that would allow brakeless robotaxi designs — a sign regulators are still actively rewriting the rulebook rather than working from settled standards.

What is riding a robotaxi actually like in 2026?

Riding a Waymo or WeRide vehicle today feels close to a normal ride-hail trip with a quieter, more cautious driving style: the car brakes earlier, gives wider berth to pedestrians and cyclists, and sometimes pauses longer than a human would at ambiguous intersections. Tesla's experience varies more by city, since fleet size is smaller and a front-seat monitor is often present, which changes both the interior feel and the trip's practical wait time.

The operators that feel most "solved" right now are the ones with years of lidar-based mapping in specific cities — Waymo in Phoenix, Baidu in Wuhan — where the car has effectively memorized the terrain. Newer markets and newer operators still show more hesitation and longer pickup windows, which matters more to a daily rider than any brand's safety statistics.

Robotaxi operators compared (August 2026)

Frequently Asked Questions

Is Tesla's Robotaxi actually driverless in 2026?

Not consistently. Tesla launched its Robotaxi service in Austin in mid-2025 with a human safety monitor riding in the front passenger seat, and as of August 2026 many trips across its roughly seven metro areas still include that monitor, unlike Waymo's or WeRide's no-human-onboard model in their driverless markets.

Which robotaxi company operates in the most cities?

Baidu's Apollo Go operates in the most cities overall, covering more than 27 Chinese cities as of 2026 plus expansion into Dubai, Abu Dhabi, and Seoul. Among US-based operators, Waymo leads with nine driverless metro areas after its February 2026 expansion into Dallas, Houston, San Antonio, and Orlando.

Is it cheaper to take a robotaxi or Uber in 2026?

It depends on the operator. Tesla's Robotaxi averages about $8.17 per ride in the SF Bay Area, roughly half of Uber or Lyft, while Waymo averages about $19.69 versus Uber's $17.47 — a 13% premium that has shrunk from 30–40% in 2025, per TechCrunch's January 2026 pricing analysis.

Have robotaxis caused any serious accidents in 2026?

Yes, though most disclosed incidents have been minor. Tesla reported 17 crash narratives to NHTSA between July 2025 and March 2026, mostly property damage with one hospitalization-level injury, and Waymo is under an active NHTSA probe after a January 2026 incident in Santa Monica in which a vehicle struck a 9-year-old child near a school.