Does Waymo still need Uber?

When can Waymo offer its own app in Austin and Atlanta?

Waymo can add its own app in Austin and Atlanta from January 2028, according to Uber, and the current Uber contract in those cities ends in May 2028.

Those dates came out in July 2026, after the Financial Times reported that Waymo had discussed leaving the partnership. The report described practical disputes. Waymo complained about dirty vehicles and poor fleet routing. Uber criticized Waymo cars that became unavailable during bad weather and reportedly called the financial terms unsustainable. Uber then told TechCrunch that Waymo had informed it of the January 2028 plan, to run alongside Uber, and that the contract ends in May 2028.

Neither company has announced a breakup. Waymo’s own site still says its Austin and Atlanta service is available only on the Uber app. So for now, nothing has changed for riders in those two cities.

The friction runs deeper than a contract renewal. Uber wants autonomous cars that behave like one more ride category inside its marketplace. Waymo wants control over how its cars, technology and brand reach passengers. Those goals fit during the early launch years. They pull apart as Waymo gets big enough to run cities alone.

Phoenix already showed the direction. The companies ended their small Phoenix arrangement earlier in 2026, and Waymo moved those cars back into its own fleet. The relationship also has a public edge: Uber’s technology chief posted a video of Waymo driving he called unsafe, and the two companies sit on opposite sides of several policy fights.

Waymo still depends on Uber in Austin and Atlanta. It now has a dated exit path, which tells you it sees that dependence as temporary.

What would it actually mean for Waymo to need Uber?

Waymo truly needs Uber only if losing the partnership would stop it from finding riders, operating vehicles, entering cities or funding a viable business.

Uber can be very useful without being essential. A useful partner makes a service faster, cheaper or easier to scale. An essential partner controls something the company cannot realistically replace.

For Waymo, the questions are simple. Can its own app keep cars busy? Can another company clean, charge and maintain the fleet? Can it open cities without Uber? And can it afford the extra work of dealing directly with passengers?

The evidence answers the first three with a clear yes. Waymo opened new cities through its own app throughout 2026, and Lyft’s Flexdrive took over fleet work in Nashville. Funding is also no longer in doubt after the $5 billion loan in October 2026. What stays hidden is profit: neither Waymo nor Uber publishes the revenue split, operating costs or margin per autonomous trip.

Where Uber could be essential	What would prove dependence	What the evidence currently shows
Passenger demand	Waymo cars remain underused without Uber bookings	Waymo draws large direct demand in most of its cities, and now adds Lyft bookings in Nashville
Fleet operations	No other company can manage charging, cleaning and depots	Moove, Avis and Lyft’s Flexdrive already perform these jobs
Market expansion	Waymo cannot enter cities through its own app	13 of Waymo’s 15 US public-service cities use the Waymo app
Economics	Direct operations cost far more than using Uber	Public data remain insufficient; financing is in place

How much of Waymo still depends on Uber today?

Waymo relies on Uber in two of the 15 US cities where it now offers public rides, so Uber controls a meaningful slice of the network, not the network itself.

Waymo said Las Vegas became its 15th US commercial city when public rides began there in September 2026. Thirteen of those cities use the Waymo app: the San Francisco Bay Area, Los Angeles, Phoenix, Miami, Orlando, Nashville, Dallas, Houston, San Antonio, Denver, San Diego, Tampa and Las Vegas. Austin and Atlanta remain Uber-only. Waymo’s rides page counts slightly differently and lists Las Vegas as coming next, because riders there are still being let in gradually from a waitlist.

That puts about 13% of Waymo’s city footprint behind Uber, down from 18% when Waymo had eleven markets. The share of rides could be higher. Austin and Atlanta have hundreds of vehicles and are more mature than Denver or Las Vegas. Neither company publishes rides by city, so city count says nothing precise about bookings, revenue or fleet.

The latest ride figure Waymo has disclosed remains more than 500,000 fully autonomous paid rides per week, disclosed in early 2026. No newer weekly total has been published. Semafor reported in October 2026 that the fleet stands at about 4,000 vehicles, and local coverage of the Las Vegas launch also put it above 4,000.

Nashville adds a nuance. Riders there can now use Lyft as well as the Waymo app, but that is extra distribution, not dependence. Waymo still sells rides directly in that city.

Waymo booking model	Current markets	Share of current city footprint
Waymo app	San Francisco Bay Area, Los Angeles, Phoenix, Miami, Orlando, Nashville (also on Lyft), Dallas, Houston, San Antonio, Denver, San Diego, Tampa and Las Vegas	13 of 15 markets
Uber app	Austin and Atlanta	2 of 15 markets

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Did Uber make Waymo’s Austin and Atlanta launches materially stronger?

Yes. Uber strengthened Waymo’s Austin and Atlanta launches by supplying instant demand, local fleet operations and a backup network of human drivers.

Riders in those cities did not need to discover Waymo, download another app or decide to choose a robotaxi. They ordered an UberX, Uber Comfort or another eligible product and were matched with a Waymo when one was nearby. Autonomous cars landed inside a booking habit people already had.

Uber also handled dispatch, cleaning, repairs and depots. The partnership was designed to reach hundreds of vehicles, which let Waymo enter both cities without building every local function itself.

An estimate from Wolfe Research gives a sense of the effect. The firm calculated that Waymo cars in Austin completed around 28 trips per day, against roughly 21 to 22 in San Francisco. That would make Austin utilization about 30% higher, even though San Francisco is Waymo’s most mature direct market.

This is a model, not official Waymo data, and the cities differ. We found no newer comparison of Austin and Atlanta with and without Uber. Even so, Uber clearly solved the early demand problem. A new car entering Austin could draw bookings from a large pool of existing ride-hail customers on its first day.

Uber gave Waymo a stronger launch engine. Thirteen direct-app cities show that the engine can be replaced.

Can Waymo attract enough riders through its own app?

Yes. Waymo now attracts large demand through its own app, and 2026 turned that into a pattern rather than a San Francisco exception.

The openings came one after another. Waymo opened its app to everyone in Dallas on 4 August 2026 and in Houston on 20 August. On 14 October, it opened San Antonio, San Diego and Tampa to all riders, with service at San Antonio’s airport. Earlier in the year, more than 150,000 people from the waitlists had ridden in Miami and Orlando before those cities opened to everyone.

Las Vegas shows the same pull before launch. Local reporting said more than 100,000 people joined the waitlist before public rides began in September 2026. Waitlists do not prove that curious users become regulars. They do show that Waymo can build a large opening audience without hiding inside Uber.

Waymo is also adding features a novelty service would not need. Waymo Premier costs $29.99 per month and offers priority matching, 10% back in Waymo credits, early access and easier cancellations, starting in San Francisco, Los Angeles and Phoenix. In Nashville, teenagers aged 13 to 17 can now ride on their own with a parent’s permission, and airport pickups began at Nashville International in October.

A paid membership targets commuters and frequent riders. It also gives Waymo direct data on frequency, loyalty and feedback, which Uber would hold if the booking went through its app.

Retention rates, acquisition costs and trips per active customer are still not disclosed. That keeps us from calling the direct demand engine mature. But the evidence now looks like a real consumer platform: waitlists, open-to-all service in about ten cities, airport service and a subscription.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Has Waymo proved it can compete directly with Uber?

Yes, in San Francisco. Waymo has shown that a direct robotaxi service can take real ride-hail share from Uber, though the hard evidence still comes from that one city.

Wolfe Research estimated that Waymo captured about 22% of rides that start and end inside its San Francisco zone. That put Waymo within about one percentage point of Lyft, while Uber kept around 55%.

These are estimates, not official statistics. Wolfe combined regulatory data with its own modelling, and the figure excludes trips outside Waymo’s zone. We found no newer share estimate, so these numbers describe the first half of 2026 at best, not today.

The scale of the shift is still hard to dismiss. Waymo had almost no share when it started selling rides in San Francisco. Reaching about one-fifth of eligible local rides in roughly two years means people use it as transport, not as a one-time demo.

Prices have moved closer too. Obi compared 94,348 identical ride requests across Waymo, Uber, Lyft and Tesla in the Bay Area. The average Waymo fare was $19.69 against $17.47 for Uber, about 12.7% more. Months earlier, the gap had been 30% to 40%. On trips of 4.3 to 9.3 kilometers, Waymo was only 2% more expensive per kilometer, and its average pickup estimate of 5.74 minutes often beat Uber outside the afternoon peak.

Waymo may be subsidizing these fares while it grows, and nothing public shows profit per trip. Still, the direct service is close enough on price and wait time to compete for ordinary journeys. Nashville adds a different proof: Waymo now competes for riders in its own app while also feeding them through Lyft, and it keeps both.

Does Uber’s customer base still give Waymo a decisive advantage?

Uber’s customer base still gives Waymo a strong launch advantage, but it is no longer the only large app Waymo can plug into.

Uber reported 208 million monthly active platform consumers and 3.867 billion trips in the second quarter of 2026, up 16% and 18% from a year earlier. Those figures mix rides and food delivery across many countries, so they overstate the pool Waymo could serve. They still explain Uber’s bargaining power.

For scale, Waymo’s last disclosed run rate of 500,000 weekly rides would produce about 6.5 million rides in a thirteen-week quarter. Uber’s quarterly trip count is roughly 600 times larger.

The real advantage is density inside each city. Uber gets requests at almost every hour, in most neighborhoods, for several vehicle types. Adding a few hundred robotaxis to that flow gives Uber many chances to find a nearby rider and cut empty time.

Waymo must build that density inside its own service area. That is hard at launch and easier once the brand is familiar. San Francisco shows the end point. Nashville shows a middle path: a Lyft account can now feed the same Waymo fleet, which means Uber’s audience is valuable but not unique.

Uber’s audience matters most in a city’s first months. Its weight falls as Waymo’s local demand grows and other apps compete to offer Waymo rides.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Does Waymo still need Uber to run the physical fleet?

No. Waymo still needs companies to clean, charge, inspect and maintain its robotaxis, but Uber has been replaced in those roles in several markets.

Moove manages vehicle operations, charging and facilities for Waymo in Phoenix and Miami. Riders book through the Waymo app.

Avis provides fleet management in Dallas, including vehicle readiness, maintenance and depots. Again, the customer books through Waymo.

Lyft’s Flexdrive runs the Nashville fleet. On 2 October 2026, it opened an 80,000-square-foot depot with dozens of fast chargers, high-voltage maintenance bays and room to support hundreds of autonomous vehicles. Waymo cars began moving in on 12 October. Lyft said the site adds more than 70 full-time jobs this year. The depot proves the capacity, not the number of cars on Nashville roads today, which neither company discloses.

Uber combines these jobs in Austin and Atlanta. One partner handling demand, dispatch and daily vehicle care is convenient. Waymo’s other cities show that each job can be bought separately.

That takes more coordination, but it gives Waymo more freedom. A problem with one operator can be fixed locally without handing over the customer network or the driving technology.

Waymo depends on an operating ecosystem. No single fleet operator looks indispensable.

What does Uber offer Waymo that fleet managers cannot?

Uber offers a combination of passenger demand, fleet coordination and human-driver backup that Moove or Avis cannot match alone. Lyft now offers a similar package in Nashville.

Uber sells this package as Uber Autonomous Solutions: demand generation, dispatch, fleet support, customer service, remote assistance, financing help and local operating knowledge. Its CEO described the goal in August 2026 as building the world’s largest platform for autonomous vehicles.

The human-driver network is the hardest part for Waymo to copy. When autonomous cars are unavailable, Uber can give the customer another car. Waymo’s own app cannot replace a robotaxi with a human driver.

Disruptions show why this matters. In May 2026, Waymo paused service in four cities, including San Antonio, after vehicles drove into flooded roads. A Waymo app customer had to find another ride. An Uber customer could still be matched with a human driver.

The mixed network also helps when demand spikes. A concert, an airport delay or a sudden storm can overwhelm a fixed robotaxi fleet. Uber can shift bookings to human drivers instead of leaving the rider stranded.

Nashville changes the picture slightly. Lyft riders there can be matched with a Waymo or with a human driver in the same app, and Lyft and Waymo dispatch from one shared fleet. The fallback is no longer something only Uber can provide.

The fallback improves reliability. It is not a reason Waymo must stay with Uber, because a robotaxi-only app already works in 13 cities and a rival marketplace now offers the same backup.

Are the Waymo-Uber economics still attractive?

Nobody outside the two companies can say. Neither discloses the revenue split, the cost per ride or a renegotiated contract.

The Financial Times reported in July 2026 that Uber saw the financial terms as unsustainable. That was reporting based on unnamed sources, not a disclosed number. The same report described disputes over vehicle availability, routing and cleanliness, which suggests each side thinks it supplies the more valuable half.

Waymo supplies the driving system, expensive vehicles and much of the safety responsibility. Uber supplies bookings, local fleet work and customer support. Both carry real costs, and each has a case for a large share of the fare.

Moving bookings into the Waymo app would let Waymo keep more of each fare and control pricing, promotions, subscriptions and data. It also adds costs: marketing, payments, support, refunds and the risk of empty cars while local demand builds.

Using Uber could still be cheaper in some cities, even after Uber’s cut. No public evidence lets anyone calculate the difference.

What has changed is Waymo’s capacity to absorb the cost. It raised $16 billion in equity at a $126 billion post-money valuation earlier in 2026, then closed a $5 billion term loan in October from lenders led by PIMCO, Blackstone and Sixth Street. A loan is not a profit signal. It does show that lenders are willing to back a business that runs mostly through its own app.

Financial independence remains unproven. Strategic independence is fully funded.

Why would Waymo leave Uber when Uber still helps?

Waymo would leave Uber to avoid becoming one interchangeable robotaxi supplier inside a marketplace controlled by someone else.

When an Uber customer gets a Waymo, Uber controls the start of the transaction. Uber decides how visible autonomous rides are, what alternatives sit next to them and which autonomous company gets the booking. The rider may remember ordering an Uber more than riding in a Waymo.

Direct bookings give Waymo the whole relationship. It sets the price, presents the service, sells Premier memberships and learns who comes back. That matters more as robotaxis become a daily habit rather than a novelty.

Uber’s growing autonomous portfolio makes the issue sharper. Uber works with Zoox, Rivian, Nuro, Lucid, Motional, Avride, Wayve and others. In August 2026, Pony.ai and Uber expanded their deal to more than 2,000 planned robotaxis across five European cities. Uber’s strategy is many fleets, not Waymo as the permanent default.

Waymo has the opposite incentive. Its edge is the strongest proven driverless system at commercial scale. Letting Uber treat it as interchangeable inventory would weaken the Waymo brand and hand Uber more pricing power.

Nashville shows the alternative Waymo prefers. Since September 2026, Lyft can sell Waymo rides there, but the Waymo app keeps selling them too. Waymo gains bookings without giving one marketplace exclusive control. The January 2028 plan for Austin and Atlanta fits the same logic: Uber builds local familiarity first, then riders get a direct option.

Some customers will stay with Uber for the fallback or for keeping every trip in one app. Waymo does not need everyone. It needs enough direct riders to keep its fleet busy and protect control of the product.

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Could Waymo reach its reported 20,000-robotaxi goal without Uber?

Probably yes, if the reported target is real, though a split would make some launches slower and riskier, especially in cities where few people know the Waymo brand.

The target itself is reported, not confirmed. Semafor reported in October 2026 that Waymo had about 4,000 vehicles and aimed for 8,000 by the end of 2026 and 20,000 by the end of 2027. Waymo did not comment. Its own loan announcement mentioned expansion in the US and abroad but gave no fleet number. Planned cars are not an operating fleet.

If the plan is real, the hard part is physical, not commercial. Five times more cars means more depots, chargers, technicians and cleaning crews. Uber removes some of that work in two cities. Moove, Avis and Flexdrive already do it elsewhere, and Waymo builds its Ojai vehicles with Magna.

The 2026 rollout shows Waymo can open cities in batches without Uber. It launched Dallas, Houston, San Antonio and Orlando together early in the year. It invited riders in Denver, San Diego and Tampa on 1 September 2026, then Las Vegas on 14 September, and opened San Antonio, San Diego and Tampa to everyone on 14 October. The newest cities run mostly on the Ojai vehicle and its sixth-generation Waymo Driver.

Waymo is also preparing launches outside the US on its own terms: Tokyo with Nihon Kotsu and GO in 2027, first passenger rides in Madrid in 2027 and Toronto in 2028. These are plans, not live service.

Fast growth can expose weak spots that mature markets hide, as the May 2026 flood pauses showed. Uber would reduce that risk in a new city.

The cost of leaving Uber is slower and less predictable expansion in a few places. With $21 billion in new equity and debt in 2026 and partners for fleet work, Waymo can accept that cost.

Could Uber replace Waymo with other robotaxi partners?

Not today. Uber has no other partner running driverless rides at Waymo’s scale in the US, though its pipeline could change the balance after 2027.

Waymo completes hundreds of thousands of paid rides a week without safety drivers. Most rival programs on Uber run in small zones, use safety operators, test without passengers or target launches years away.

Zoox is one of the closest alternatives. On 30 July 2026, US regulators granted it a temporary exemption to deploy up to 2,500 vehicles per year for two years. That is a ceiling, not a fleet. Uber plans to offer Zoox rides in Las Vegas and later Los Angeles, while Zoox keeps its own app.

Abroad, Pony.ai and Uber plan more than 2,000 robotaxis across five European cities, starting from Zagreb. No delivery dates have been given for the other four.

Uber’s Rivian deal could become the largest. Uber or its fleet partners expect to buy 10,000 autonomous R2 vehicles, with an option for 40,000 more, with first deployments planned for 2028. The Lucid and Nuro program is testing on public roads, with commercial service planned and Houston targeted for 2027. None of this is existing driverless supply.

Uber is building credible alternatives, but Waymo remains its largest proven US partner. Over time, the portfolio could turn Waymo from a scarce supplier into one option among several. That is exactly why Waymo wants its own channel.

Uber robotaxi partner	Current position	How close it is to replacing Waymo
Waymo	Large-scale fully autonomous service in 15 US cities, two of them Uber-only	Uber’s most proven autonomous supplier
Zoox	Limited commercial service; federal exemption for up to 2,500 vehicles per year from July 2026	Credible technology with a much smaller footprint
Pony.ai	More than 2,000 robotaxis planned across five European cities	Outside the US; mainly future capacity
Rivian	Up to 50,000 vehicles contemplated, starting with deployments planned for 2028	Potentially large, but mainly future capacity
Lucid and Nuro	Autonomous testing underway, with commercial launches planned	Serious program that has not reached Waymo’s operating scale

If you want more recent data on this point, please see our latest autonomous vehicle market report.

Could Waymo work with Uber the way it now works with Lyft in Nashville?

Yes. Nashville is a working template for a looser deal, and it could fit Austin and Atlanta once Uber’s exclusivity ends.

Since 9 September 2026, Lyft riders in central Nashville can be matched with a Waymo at no extra cost when they order a standard or comfort ride. The trip must start and end inside the service area. Riders can decline a robotaxi, and human drivers keep serving the city. The Waymo app sells rides from the same fleet. Flexdrive, Lyft’s subsidiary, runs the cars.

A future Waymo-Uber deal could copy that structure. Waymo would keep direct booking as its main channel while letting Uber customers reach some cars during launches, quiet hours or specific trip types. Uber would get autonomous supply without being Waymo’s only route to riders. That is close to what Uber says Waymo already told it: its own app from January 2028, alongside Uber.

This would ease the fight over control. Waymo keeps subscriptions, pricing and customer data. Uber keeps offering a mixed network of human and autonomous cars.

The disputes reported in July 2026 may make that compromise harder in Austin and Atlanta. Commercial incentives often outlast operational tension, though, when each side still holds something the other wants.

Waymo and Uber no longer need a permanent, exclusive alliance. A Nashville-style arrangement could still pay off for both.

Does Waymo still need Uber?

No. Waymo no longer needs Uber to build a viable robotaxi business. It needs Uber only in Austin and Atlanta, and only until its contract lets it sell rides directly.

Waymo’s own app serves 13 of its 15 US public-service cities. In 2026 alone, Waymo opened Dallas, Houston, San Antonio, San Diego and Tampa to everyone and began rides in Denver and Las Vegas. It runs a paid membership, has approached Lyft’s estimated share in part of San Francisco and has brought prices close to Uber’s. Moove, Avis and Flexdrive show that fleet work can be outsourced without giving up the passenger.

Uber keeps two real advantages. Its booking network can fill new cars immediately, and its human drivers can complete trips when Waymo pauses. Those strengths make Uber an excellent launch and reliability partner. Nashville shows that Lyft can now offer much of the same.

The open question is cost. Nobody outside the companies knows whether Waymo earns more on its own, because the partnership economics and Waymo’s profit per ride stay private. The $5 billion loan proves Waymo can fund independence, not that independence pays better.

The direction is hard to miss. Waymo ended the Phoenix arrangement, built several fleet partnerships, launched a non-exclusive deal with Lyft and, according to Uber, plans its own app in Austin and Atlanta from January 2028. A reported plan to grow to 20,000 vehicles by the end of 2027, if confirmed, would make direct control even more valuable.

Uber has gone from potential gatekeeper to one distribution channel among several. Waymo might grow more slowly without Uber in a few cities, but Uber no longer decides whether Waymo grows.

If you want more recent data on this point, please see our latest autonomous vehicle market report.