What are the top autonomous vehicle startups by revenue?

Everything we know about the autonomous vehicle market, in one place. Get our report whether you build or invest.
SUMMARY
Waymo is currently the strongest candidate for the largest autonomous vehicle startup by revenue scale, while WeRide and Pony.ai have the largest clean full-year revenue figures we can verify directly.
The ranking is less straightforward than it looks. Waymo's roughly $355 million annualized revenue is an external estimate, while WeRide's $97.9 million and Pony.ai's $90 million are reported company revenue, so the top three do not rest on the same kind of evidence.
The market drops off quickly after those leaders. Inceptio Technology, Trunk Technology, TIER IV and KargoBot cluster around roughly $40 million to $60 million of annual revenue, which shows how rare nine-figure autonomous-driving businesses still are.
Autonomous trucking has produced a broader middle tier than robotaxis. Trunk Technology, KargoBot and Inceptio already operate at tens of millions of dollars in annual revenue, while Gatik has more than $600 million of contracted revenue even though that should not be confused with one year's recognized sales.
China currently has the deepest bench of autonomous vehicle startups with visible revenue. WeRide and Pony.ai lead, but the more interesting point is that several freight and systems companies below them are already monetizing at meaningful scale.
Robotaxi economics are finally becoming visible inside public-company results. Pony.ai generated $12.1 million of robotaxi-services revenue in Q2 2026, while WeRide's broader L4 businesses reached $18.5 million in the same quarter.
Fleet size alone is a weak revenue proxy. Pony.ai and WeRide operate large robotaxi fleets, but Kodiak shows that a much smaller number of customer-owned driverless trucks can still generate millions of dollars when the vehicles are tied directly to paid industrial work.
Software and deployment services matter more than the consumer-facing narrative suggests. TIER IV generated ¥6.41 billion in annual revenue without building its business around a single ride-hailing network, showing that suppliers can monetize years before full autonomy becomes ubiquitous.
The sector is commercializing faster than it is becoming profitable. WeRide, Pony.ai, May Mobility and TIER IV all show that meaningful revenue can coexist with very large operating losses and cash burn.
The cleanest conclusion is that autonomous vehicle revenue is real but still highly concentrated. A handful of companies now operate at serious commercial scale; far fewer have crossed $100 million of annual revenue, and almost none have demonstrated mature company-level economics.

Failures teach as much as successes. We break down the biggest falls in our autonomous vehicle market report.
The ranking of top startups in the autonomous vehicle market by revenue
Below is a table ranking all the companies in this market by their current revenue scale. You can find our methodology at the end of this page.
If you want a deeper understanding of the market and its current dynamics, get our report covering the Autonomous Vehicle Market.
| Ranking | Company | Latest Metric | Metric Type | Freshness | Disclosed When | Source Quality | Confidence | Segment | Why This Ranking |
|---|---|---|---|---|---|---|---|---|---|
| 1 | Waymo | $355M | Revenue Run-Rate | Fresh · 7mo | May 5, 2026 | Third-Party Estimate | Medium | Robotaxi | Largest credible current revenue-scale figure found. The estimate is weaker than audited filings below, but the gap versus ~$90–100M annual revenue at listed AV peers is substantial. |
| 2 | WeRide | $97.9M / Rmb685M | Fiscal-Year Revenue | Fresh · 9mo | Mar 2026 | Filed / Audited | Medium | Robotaxi / Robobus / Robovan | Below Waymo because Waymo's current estimated run-rate is much larger; above Pony.ai on comparable audited FY2025 revenue. H1 2026 results also support continued scale. |
| 3 | Pony.ai | $90M | Fiscal-Year Revenue | Fresh · 9mo | Mar 26, 2026 | Filed / Audited | High | Robotaxi / Robotruck | Slightly below WeRide on comparable FY2025 revenue. Very fresh 2026 quarterly results show acceleration and support its position above the next tier. |
| 4 | Trunk Technology | Rmb345M | Fiscal-Year Revenue | Fresh · 9mo | Jun 12, 2026 | Filed / Audited | High | L4 Autonomous Freight | Strong recent full-year revenue with a relatively clean L4 trucking perimeter, putting it ahead of similarly sized but less comparable evidence below. |
| 5 | TIER IV | ¥6.41B | Fiscal-Year Revenue | Fresh · 12mo | Jun 29, 2026 | Filed / Audited | High | Autonomous Driving Software / Shuttles | Exceptionally strong source quality and meaningful scale; placed just below Trunk because Trunk's disclosure is a little more directly comparable to commercial L4 trucking revenue. |
| 6 | KargoBot | Rmb300M | Annual Revenue | Aging · 20mo | Jan 24, 2025 | Company Disclosed | Medium | L4 Autonomous Freight | Clean L4 trucking revenue deserves more weight than Inceptio's larger but older third-party figure with a mixed L2–L4 perimeter. |
| 7 | Inceptio Technology | ~Rmb430M | Annual Revenue | Aging · 21mo | Dec 11, 2025 | Third-Party Estimate | Low | Autonomous Trucking | Nominally larger than KargoBot, but the figure is older, estimated and includes a materially less clean SAE-level perimeter. |
| 8 | Gatik | >$600M | Bookings | Very Fresh · 1mo | Aug 25, 2026 | Company Disclosed | Medium | L4 Autonomous Freight | The largest financial commercial signal after Waymo, but $600M is contracted multi-year business rather than recognized revenue, so it stays below companies with sizable actual annual revenue. |
| 9 | Autonomous A2Z | Krw16B | Annual Revenue | Fresh · 9mo | Jan 29, 2026 | Company Disclosed | Medium | L4 Shuttle / Robotaxi | Recent company-wide revenue is stronger than the smaller pure-play figures below, but its pre-full-commercialization project mix limits comparability. |
| 10 | May Mobility | ~$10M | Annual Revenue | Fresh · 9mo | Sep 16, 2026 | Filed / Audited | High | Robotaxi / Autonomous Shuttle | Smaller than A2Z, but unusually clean and current revenue evidence from a pure-play autonomous-mobility company. |
| 11 | Kodiak AI | $3.5M | Quarterly Revenue | Very Fresh · 2mo | Aug 7, 2026 | Filed / Audited | High | Autonomous Trucking | Direct, current SEC-reported revenue gets substantial weight despite the smaller amount and ranks ahead of mixed-perimeter or less-authoritative financial evidence. |
| 12 | Einride | ~$27M (Sek273M Reported) | Six-Month Revenue | Very Fresh · 3mo | Aug 18, 2026 | Company Disclosed | Low | Autonomous & Electric Freight | Nominal company revenue is much larger than nearby companies, but the inability to isolate the L4 component warrants a major perimeter discount. |
| 13 | Beep | $36M | Bookings | Fresh · 15mo | Jun 25, 2025 | Credible Reported | Medium | Autonomous Shuttle / Transit | Concrete contracted dollars are stronger evidence than non-financial fleet proxies, but the multi-year nature keeps it below recognized-revenue companies. |
| 14 | Aurora Innovation | $2M | Quarterly Revenue | Very Fresh · 2mo | Jul 29, 2026 | Filed / Audited | Medium | Autonomous Trucking | Smaller than Bot Auto's reported annual estimate, but the filed recognized revenue is sufficiently higher-quality to rank Aurora above it. |
| 15 | Bot Auto | $7.6M | Annual Revenue | Fresh · 9mo | Aug 10, 2026 | Third-Party Estimate | Low | L4 Autonomous Trucking | The annual estimate is meaningful, but its substantially weaker provenance places it below Aurora's smaller SEC-reported figure. |
| 16 | Oxa | £3.09M | Fiscal-Year Revenue | Aging · 21mo | Sep 25, 2025 | Filed / Audited | Low | Autonomous Driving Software / Shuttles | Audited revenue is strong evidence, but it is older and includes businesses outside the strict public-road definition. |
| 17 | Baidu (Apollo Go) | 3.2M Fully Driverless Rides | Transactions | Very Fresh · 6mo | May 18, 2026 | Company Disclosed | Medium | Robotaxi | Strongest current operating-volume signal among companies without usable segment revenue: millions of fully driverless rides in a single quarter. |
| 18 | Zelostech | 25,000+ Vehicles | Units Deployed | Very Fresh · 1mo | Aug 25, 2026 | Company Disclosed | Low | L4 Robovan / Autonomous Logistics | Exceptional deployed fleet scale, but the mixed operating perimeter is less directly comparable with Apollo Go's fully driverless ride transactions. |
| 19 | Neolix | 13,000+ Vehicles | Units Sold | Very Fresh · 0mo | Sep 2026 | Company Disclosed | Medium | L4 Robovan / Urban Delivery | More directly commercial than most non-revenue proxies because these are delivered vehicles, though no price or recognized revenue is disclosed. |
| 20 | Tesla | ~2.4M Paid Robotaxi Miles | Other Scale Signal | Very Fresh · 3mo | Jul 22, 2026 | Company Disclosed | Low | Robotaxi | Paid usage provides a genuine commercial indicator, but mixed supervised/unsupervised operation and absent segment revenue make it weaker than delivered-unit evidence above. |
| 21 | MOGOX | 20+ Cities | Locations | Very Fresh · 0mo | 2026 | Company Disclosed | Low | L4 Robobus | Multi-city operations and a real public-transit procurement validate commercial activity, but the Singapore consortium contract cannot be attributed entirely to MOGOX. |
| 22 | Avride | 60,000+ Rides | Transactions | Very Fresh · 3mo | Jun 4, 2026 | Company Disclosed | Medium | Robotaxi | A clean commercial-ride metric puts Avride above less directly monetized fleet/location proxies below. |
| 23 | Motional | 130,000+ Rides And Deliveries | Transactions | Very Fresh · 0mo | Feb 2026 | Company Disclosed | Low | Robotaxi | Higher headline transaction count than Avride, but the much longer cumulative period and current supervised phase make it weaker evidence of present commercial scale. |
| 24 | Rino | 200+ Cities | Locations | Very Fresh · 2mo | Jul 2, 2026 | Company Disclosed | Low | L4 Robovan / Logistics | Very broad footprint but a location count is less useful for revenue ranking than actual commercial transactions above. |
| 25 | Zoox | ~100 Robotaxis | Units Deployed | Very Fresh · 0mo | Sep 17, 2026 | Credible Reported | Medium | Robotaxi | Current paid service and a verified fleet give useful evidence, though operations remain much smaller than the higher-ranked networks. |
| 26 | QCraft | 100 Vehicles | Other Scale Signal | Very Fresh · 0mo | Apr 20, 2026 | Credible Reported | Low | L4 Autonomous Logistics | Concrete near-term commercial order is stronger than distant deployment commitments below, but cannot be treated as revenue. |
| 27 | Verne | Several Thousand Customer Rides | Transactions | Very Fresh · 0mo | Sep 10, 2026 | Company Disclosed | Low | Robotaxi Operator | Actual customer transactions outrank large future commitments, although the safety-operator configuration limits strict L4 comparability. |
| 28 | Nuro | 35,000+ Vehicles | Other Scale Signal | Very Fresh · 5mo | Apr 14, 2026 | Company Disclosed | Low | Robotaxi Technology | Very large commercially backed commitment, but because deployment is future-dated it ranks below companies already carrying paying customers. |
| 29 | Waabi | 25,000+ Robotaxis | Other Scale Signal | Fresh · 8mo | Jan 28, 2026 | Company Disclosed | Low | Autonomous Trucking / Robotaxi Technology | Large commitment validates commercial demand, but the program is less operational today than Nuro's already structured OEM/Uber rollout. |
| 30 | Wayve | 15 Vehicles | Units Deployed | Very Fresh · 0mo | Sep 13, 2026 | Credible Reported | Low | Robotaxi / AV Software | Actual commercial vehicles matter more than testing alone, but the small fleet and supervised operation keep Wayve below mature commercial networks. |
| 31 | DiDi Autonomous Driving | 2 Cities | Locations | Very Fresh · 1mo | Aug 31, 2026 | Company Disclosed | Low | Robotaxi | More technologically strict than some supervised programs, but too little commercial-volume information exists to rank it higher. |
| 32 | RideFlux | 112-Km Paid Freight Route | Other Scale Signal | Very Fresh · 0mo | Apr 16, 2026 | Company Disclosed | Low | Autonomous Trucking | Demonstrably commercial, but a single route is a much narrower scale signal than the networks above. |
| 33 | AutoX | 1,000+ Robotaxis | Units Deployed | Historical · 48mo | Mar 22, 2023 | Company Disclosed | Low | Robotaxi | The deployment was significant, but freshness is poor enough that much smaller current commercial signals deserve greater weight. |
| 34 | BMW | 1 Commercial L3 System | Other Scale Signal | Very Fresh · 0mo | Sep 2026 | Company Disclosed | Low | L3 Passenger Cars | Clear customer-facing L3 commercialization qualifies BMW, but absence of sales-volume data prevents a meaningful revenue-scale comparison. |
| 35 | Mercedes-Benz | 2 Model Ranges | Other Scale Signal | Aging · 21mo | Dec 17, 2024 | Company Disclosed | Low | L3 Passenger Cars | Broader historic L3 commercialization than most automakers, but no sales count exists and 2026 evidence suggests reduced emphasis on L3 in the latest refresh. |
| 36 | BAIC / ARCFOX | 3 Plated L3 Vehicles | Units Deployed | Fresh · 9mo | Dec 26, 2025 | Credible Reported | Low | L3 Passenger Cars | A precise current unit signal is preferable to mere approval, but three vehicles clearly represent only an early commercialization stage. |
| 37 | Changan Automobile | 1 L3-Approved Model | Other Scale Signal | Fresh · 9mo | Apr 3, 2026 | Credible Reported | Low | L3 Passenger Cars | Qualifies technically, but evidence of actual commercial scale is weaker than BAIC's disclosed plated units. |
| 38 | Honda | 100-Unit Limited-Production Plan | Other Scale Signal | Historical · 60mo | Mar 4, 2021 | Company Disclosed | Low | L3 Passenger Cars | Honda provides clean historical evidence of L3 commercialization, but the figure is too old to say much about current scale and therefore ranks last. |

Not every big name makes it. Our autonomous vehicle market report looks at who stumbled, why, and what the survivors did differently.
Which autonomous vehicle startups make the most revenue today?
Waymo looks like the largest autonomous vehicle startup by current revenue scale, while WeRide and Pony.ai have the strongest large annual revenue figures we can verify directly.
The distinction is important. Waymo does not publish standalone revenue, so its roughly $355 million annualized revenue figure comes from an external estimate based on its commercial activity. WeRide reported $97.9 million of 2025 revenue and Pony.ai reported $90 million, both through formal financial reporting.
Below those three, the market becomes much more crowded. Inceptio Technology appears to be around the $60 million range on older industry data, Trunk Technology generated roughly $48 million in 2025, TIER IV generated about $43 million, and KargoBot has disclosed annual revenue of roughly $42 million.
We deliberately keep Gatik separate from that group. The company currently reports more than $600 million in contracted revenue, which is a very strong commercial result but cannot be compared directly with one year of recognized revenue.
| Rank | Company | Best revenue evidence | Metric | How strong is the evidence? |
|---|---|---|---|---|
| 1 | Waymo | ~$355M | Estimated revenue run-rate | Strong scale evidence, but not company-reported revenue |
| 2 | WeRide | $97.9M | FY2025 revenue | Filed annual revenue |
| 3 | Pony.ai | $90M | FY2025 revenue | Filed annual revenue |
| 4 | Inceptio Technology | ~RMB430M / ~$60M | 2024 annual revenue estimate | Older third-party industry estimate |
| 5 | Trunk Technology | RMB345M / ~$48M | FY2025 revenue | IPO-grade financial disclosure |
| 6 | TIER IV | ¥6.41B / ~$43M | FY2025 revenue | Audited IPO financials |
| 7 | KargoBot | RMB300M / ~$42M | Annual revenue | Company-disclosed operating revenue |
| 8 | Autonomous A2Z | KRW16B / ~$11M | 2025 revenue | Company-disclosed annual revenue |
| 9 | May Mobility | ~$10M | 2025 revenue | SEC transaction disclosure |
| 10 | Bot Auto | ~$7.6M | Estimated annual revenue | Third-party estimate |
| 11 | Oxa | ~£3M | Filed annual revenue | Older UK company accounts |
| 12 | Kodiak AI | $3.5M | Q2 2026 revenue | Current SEC-reported quarterly revenue |
Is Waymo really the biggest autonomous vehicle startup if it does not report revenue?
Yes, Waymo is currently far enough ahead operationally that the available revenue estimate still puts it in a different league from other autonomous vehicle startups.
We should be precise about what we know. Alphabet does not publish a standalone Waymo income statement, so we cannot point to an audited $355 million annual revenue figure. What we can see directly is the commercial engine underneath the estimate.
Waymo said earlier this year that it was providing more than 400,000 rides per week across six major US metropolitan areas. It also said that 2025 ride volume more than tripled to 15 million rides. The service has since kept expanding into new cities.
That level of paid activity is far beyond most autonomous vehicle startups. Even allowing for uncertainty around fares, promotions and utilization, it is difficult to reconcile Waymo's operating scale with a business that would sit anywhere near the $90–100 million revenue level of WeRide and Pony.ai.
So we rank Waymo first, but with a clear label: estimated current revenue scale, not filed annual revenue.

Market share tells you more than funding headlines. See who leads in our autonomous vehicle market report.
Which autonomous vehicle startups have crossed $100 million in revenue?
Waymo is currently the only dedicated autonomous vehicle company in our research with credible evidence clearly above a $100 million annual revenue scale.
WeRide came close at $97.9 million in 2025, while Pony.ai reached $90 million. Those are the two largest directly reported annual figures we found among pure-play autonomous driving companies.
The threshold is surprisingly hard to cross. Autonomous vehicles have attracted enormous amounts of capital, yet the number of companies with visible annual revenue around $100 million remains tiny.
Gatik may look like an exception because it has signed more than $600 million of contracted business. That figure represents future contracted revenue across customer deployments rather than one year's sales, so putting Gatik into the $600 million annual-revenue bucket would badly distort the comparison.
Today, $100 million of actual annual autonomous-driving revenue is still a major commercial milestone rather than a normal outcome for a well-funded startup.
Is WeRide still bigger than Pony.ai by revenue?
WeRide remains slightly bigger on the latest completed full year, but Pony.ai moved ahead during the first half of 2026.
WeRide generated $97.9 million in 2025, up 89.6% from a year earlier. Pony.ai generated $90 million, up 20%. On that clean full-year comparison, WeRide still leads.
The latest six-month numbers flip the order. Pony.ai reached $70.5 million of revenue in the first half of 2026, while WeRide reported $51 million. Their second quarters were much closer at $36.2 million for Pony.ai and $34.2 million for WeRide.
Neither company is slowing down. WeRide's second-quarter revenue grew 82.2% year over year, while Pony.ai grew 68.8%.
For now, the best answer depends on the period. WeRide won the last completed year; Pony.ai is currently running ahead in 2026.
| Metric | WeRide | Pony.ai |
|---|---|---|
| FY2025 revenue | $97.9M | $90.0M |
| FY2025 revenue growth | +89.6% | +20.0% |
| H1 2026 revenue | $51.0M | $70.5M |
| Q2 2026 revenue | $34.2M | $36.2M |
| Q2 2026 YoY growth | +82.2% | +68.8% |
| Current robotaxi fleet | 1,800+ | 1,975 |
| Q2 robotaxi-related revenue disclosed separately? | L4 businesses reported together | $12.1M robotaxi services |
How much of WeRide's revenue actually comes from autonomous vehicles?
Most of WeRide's business is tied to autonomous-driving products and services, but its revenue is much broader than passengers paying for robotaxi rides.
WeRide generated RMB684.6 million in 2025. Product revenue accounted for RMB359.8 million, helped by sales of robotaxis, robobuses, robosweepers and robovans. Service revenue contributed another RMB324.7 million through areas including intelligent-data services and autonomous-driving technical support.
The company sold 123 robotaxis, 156 robobuses, 91 robosweepers and 28 robovans during the year.
More recently, WeRide has begun reporting its L4 businesses more explicitly. L4 revenue reached $18.5 million in Q2 2026, up 47.3% year over year, while total company revenue reached $34.2 million.
That makes WeRide quite different from a pure ride-hailing operator. A large part of the business currently comes from selling autonomous-driving technology, vehicles and services around the network itself.

Is the market overpaying? Our autonomous vehicle market report compares valuations across the players that matter.
Is Pony.ai actually making meaningful money from robotaxis now?
Yes, Pony.ai's robotaxi business has finally become a meaningful revenue line rather than a small add-on to engineering and trucking work.
Robotaxi services generated $12.1 million in Q2 2026, up 691% from a year earlier. Fare-charging revenue grew even faster, and robotaxi services reached $20.6 million for the first half.
The fleet has expanded alongside the revenue. Pony.ai had 1,975 robotaxis at the end of Q2 and says PonyPilot now has more than 1.5 million registered users.
Robotaxis still account for only part of the company. Q2 also included $13.3 million from robotruck services and $10.8 million from intelligent solutions.
Paid passenger activity is now large enough to move Pony.ai's financial results. A year ago, that was much harder to say.
Which Chinese autonomous vehicle startups make the most money?
China currently has the deepest group of autonomous vehicle startups with visible revenue, ranging from robotaxis to freight and vehicle systems.
WeRide and Pony.ai sit at the top among the companies with public financial statements. Below them, Inceptio, Trunk Technology and KargoBot show that Chinese autonomous trucking has already built a meaningful commercial base.
Trunk Technology reported RMB345 million of 2025 revenue. KargoBot has disclosed RMB300 million of annual revenue from autonomous freight. Industry data used in Inceptio's listing materials put its relevant 2024 revenue at around RMB430 million, although that figure is older and less clean than the filed numbers around it.
Autonomous A2Z is South Korean rather than Chinese, so once we keep the geographic boundary strict, there is a noticeable gap below that Chinese trucking group.
The depth is the interesting part. China does not have only one or two autonomous-driving companies making real money. It has several operating at tens of millions of dollars in annual revenue across very different parts of the market.
Are autonomous trucks making more money than robotaxis?
Robotaxis have the biggest individual business in Waymo, but autonomous trucking currently has a broader group of startups turning deployments into meaningful commercial revenue.
KargoBot, Trunk Technology and Inceptio all sit around the tens-of-millions-of-dollars level. Gatik has built a contracted backlog above $600 million. Kodiak is much smaller today but has moved into paid driverless operation and reported $3.5 million of quarterly revenue.
Truck autonomy has a fairly direct sales pitch. The customer already pays to move freight, routes can repeat every day, vehicle utilization can be high, and removing or reducing driver requirements attacks a visible operating cost.
Robotaxis have a harder consumer network to build, but the upside can be much larger once fleet density becomes high enough. Waymo already shows what that can look like.
So far, robotaxis have the revenue leader, while freight gives us more examples of smaller autonomous-driving businesses that have already found paying customers.

Who's who in autonomous vehicle, at a glance. Our report goes deeper into each player and where they stand.
Is Gatik really a $600 million autonomous trucking company?
Gatik has built more than $600 million of contracted revenue, but calling it a $600 million annual-revenue company would be wrong.
Gatik disclosed the figure alongside its latest financing and says its driverless trucks have completed more than 85,000 orders for major retail, grocery and consumer-goods customers.
Those contracts are commercially important. They show that large companies are willing to commit substantial budgets to driverless freight rather than simply run short pilots.
The accounting is different from annual revenue, though. Multi-year contracts are recognized over time and can depend on deployment schedules, service volumes and other contractual conditions.
Gatik therefore belongs near the top when we look at customer commitments. In a ranking based on current recognized revenue, companies such as WeRide, Pony.ai, Trunk Technology and KargoBot still have cleaner evidence.
How big are Trunk Technology and KargoBot today?
Trunk Technology and KargoBot have both built autonomous freight businesses around the $40–50 million annual revenue range, which puts them well beyond pilot-stage commercialization.
Trunk Technology generated RMB345 million in 2025, up from RMB134 million in 2023. That works out to roughly 60% compound annual growth over two years. Its operations span highway logistics, logistics hubs and urban transportation.
KargoBot has disclosed RMB300 million in annual revenue while operating hundreds of L4 autonomous freight vehicles. Its model includes autonomous convoy operations, where unmanned trucks can follow a human-driven lead vehicle on suitable routes.
Neither needed a giant consumer robotaxi network to get this far. They found a narrower commercial problem and started charging customers for it.
For autonomous trucking, that path currently looks easier to monetize than waiting for unrestricted driverless operation everywhere.
Can autonomous-driving software companies make serious revenue?
Yes, TIER IV shows that autonomous-driving software, engineering and deployment services can already support a business with more than ¥6 billion in annual revenue.
TIER IV generated ¥6.41 billion in its fiscal year ended in 2025, up 65.6% from ¥3.87 billion a year earlier. The company later listed on the Tokyo Stock Exchange Growth Market.
Its revenue does not depend on one robotaxi fleet. TIER IV sells mobility services, development work for manufacturers and autonomous-driving solutions built around the Autoware ecosystem.
Carmakers, municipalities and transport operators can spend on autonomous-driving development years before a mass-market driverless service reaches their city.
TIER IV therefore sits among the largest companies in our ranking even though most consumers will never book a ride directly from it.

Price is often what holds a market back. See how fast that's changing in our autonomous vehicle market report.
How much revenue does May Mobility make?
May Mobility generated about $10 million of revenue in 2025, which makes it a real commercial autonomous-mobility business but still a small one financially.
The figure became public through SEC materials connected to its planned listing. The same disclosure showed a 27% gross margin and roughly $93 million of cash burn during the year.
May has already provided hundreds of thousands of commercial autonomous rides, so the revenue number gives useful context to all those operating milestones. A large ride count does not automatically turn into a large top line.
Its model also relies heavily on partnerships with cities, transit systems and fleet partners rather than simply charging individual riders a standard Uber-like fare.
May Mobility has clearly moved beyond experimentation. The harder step now is getting from roughly $10 million to the revenue scale already reached by the market leaders.
Is Kodiak AI already making real autonomous-trucking revenue?
Yes, Kodiak AI is making real revenue now, although the business remains early compared with the largest autonomous-trucking companies.
Kodiak reported $3.5 million of Q2 2026 revenue, up 91% from the previous quarter. The company had 35 customer-owned driverless trucks at quarter-end and had passed 40,000 cumulative hours of paid driverless operation.
That customer-owned model is worth watching. Kodiak can sell its autonomous-driving technology and services without buying every truck itself, which could eventually make fleet expansion less capital-intensive.
The company has also delivered more than 20,000 loads.
Those figures do not put Kodiak anywhere near the revenue leaders yet. They do show that its autonomous trucks have crossed the line from demonstration to recurring paid work.
What type of autonomous vehicle startup makes the most money?
The autonomous vehicle startups making the most money today usually combine several revenue streams instead of relying on one simple business model.
Waymo is closest to a pure transportation network at the top of the market. WeRide combines vehicle sales with autonomous-driving and data services. Pony.ai makes money from robotaxis, robotrucks and intelligent-driving solutions. TIER IV sells software, engineering and deployment work. Autonomous freight companies can charge directly for transport or sell technology into customer-owned fleets.
Even Einride shows how blurry the category can become. The company reported $27 million of first-half 2026 revenue across its electric, digital and autonomous freight platform, but we cannot responsibly treat all $27 million as autonomous-driving revenue.
The commercial winners these days are finding several ways to get paid while autonomy itself is still scaling.
| Business model | Companies using it | What customers pay for |
|---|---|---|
| Robotaxi network | Waymo | Paid passenger rides |
| Vehicles + AV services | WeRide | Vehicles, support, data and autonomous operations |
| Robotaxi + trucking + systems | Pony.ai | Rides, freight and intelligent-driving products |
| Autonomous freight operations | KargoBot, Gatik | Freight movement and logistics capacity |
| AV software + engineering | TIER IV | Development, software and deployment services |
| Driver-as-a-Service | Kodiak | Autonomous driving technology on customer fleets |
| Mixed electric/autonomous freight | Einride | Transport, software, charging and autonomy |

Does the funding match reality? Our autonomous vehicle market report looks at the data behind the deals.
Does having more autonomous vehicles mean a startup makes more revenue?
No, fleet size can tell us whether a company is scaling physically, but it is a poor shortcut for revenue.
Pony.ai had 1,975 robotaxis at the end of Q2, while WeRide reported more than 1,800 robotaxis inside an L4 fleet of roughly 3,400 vehicles. Both companies now generate tens of millions of dollars each quarter, so fleet growth and revenue are clearly moving together there.
Kodiak gives us a very different model. It had only 35 customer-owned driverless trucks at quarter-end yet generated millions of dollars during the quarter because those vehicles are tied directly to paid industrial work.
The opposite can also happen. A large test fleet may collect huge mileage with little associated revenue if the vehicles are running demonstrations, subsidized programs or pre-commercial trials.
A vehicle count is useful operational information. Revenue tells us whether someone is actually paying enough for the deployment to build a business around it.
Why isn't Baidu Apollo Go near the top of the autonomous vehicle revenue ranking?
Apollo Go is one of the world's largest robotaxi operations, but Baidu does not disclose a clean standalone Apollo Go revenue figure.
The operating scale is huge. Baidu reported 3.2 million fully driverless rides in Q1 2026, more than 120% growth year over year, with weekly rides peaking above 350,000. Apollo Go had passed 22 million cumulative public rides shortly afterward.
More recently, Baidu said the service had expanded to 28 cities and its fleet had accumulated more than 240 million fully driverless kilometers.
We could try to multiply rides by an assumed fare, but that would create a revenue number Baidu never reported. Pricing differs by city, promotions matter and some deployments operate under different commercial arrangements.
So Apollo Go stays out of the numerical revenue leaderboard despite being one of the largest autonomous ride-hailing networks in the world. Keeping it there would require us to invent the number we are supposedly ranking.
Why aren't Tesla, Mercedes-Benz and BMW in this autonomous vehicle startup ranking?
Tesla, Mercedes-Benz and BMW are established automakers, and their company revenue tells us almost nothing about the commercial size of their autonomous-driving businesses.
Putting Tesla's total automotive revenue beside Pony.ai's robotaxi and robotruck revenue would create a huge number with almost no analytical value.
The same issue applies to Mercedes-Benz DRIVE PILOT and BMW's Level 3 systems. They prove that customers can buy vehicles with advanced automated-driving features, but neither company gives us a clean standalone revenue figure for that capability.
Our ranking therefore focuses primarily on autonomous-driving businesses where the commercial evidence can reasonably be connected to the technology we are studying.
Waymo remains an unusual case because it sits inside Alphabet but operates as a clearly identifiable standalone autonomous-mobility business.
Big promises are easy to make. Our autonomous vehicle market report checks which ones were actually kept.
How concentrated is autonomous vehicle startup revenue today?
Autonomous vehicle revenue remains highly concentrated, with a steep drop from the small group of leaders to everyone else.
As seen above, Waymo appears to operate at a few hundred million dollars of annualized revenue scale. The next two companies sit around $100 million of annual revenue, followed by a second group around $40–60 million.
Then the numbers fall quickly. Autonomous A2Z reported KRW16 billion in 2025, roughly $11 million. May Mobility sits around $10 million. Kodiak's latest quarter was $3.5 million.
That shape tells us more than the headline market size. Autonomous driving now has genuine commercial companies, but it still does not have dozens of startups generating hundreds of millions of dollars each.
Most of the sector remains in the gap between a technically working product and a large business.
Are the biggest autonomous vehicle startups profitable yet?
No, the leading autonomous vehicle startups are generating much more revenue than a few years ago, but losses are still enormous relative to sales.
WeRide reported an operating loss of RMB1.85 billion in 2025. Pony.ai's Q2 operating expenses alone reached $72.1 million against $36.2 million of quarterly revenue, while its net loss was $45.4 million.
May Mobility's disclosure is even more striking at a smaller scale: roughly $10 million of annual revenue alongside about $93 million of cash burn.
TIER IV also remains deeply loss-making despite rapid revenue growth. Its fiscal-year ordinary loss was ¥5.5 billion against ¥6.41 billion of revenue.
There are improvements around the edges. Pony.ai's gross margin rose to 17.5% in Q2, and several companies are getting more revenue out of each deployment. Still, the current leaders are nowhere near proving that large autonomous fleets will produce attractive company-level profits.
Which autonomous vehicle startups are growing revenue fastest right now?
WeRide and Pony.ai currently give us the clearest high-quality evidence of rapid growth at meaningful scale, while several smaller companies are also expanding quickly from much lower bases.
WeRide grew full-year revenue 89.6% in 2025 and followed that with 82.2% year-over-year growth in Q2 2026. Pony.ai's overall Q2 revenue grew 68.8%, with robotaxi services growing much faster.
Trunk Technology has also expanded sharply, going from RMB134 million of revenue in 2023 to RMB345 million two years later.
TIER IV increased fiscal-year revenue by 65.6%, while Autonomous A2Z says its 2025 revenue reached KRW16 billion, around 50% above the previous year.
Kodiak's 91% quarter-over-quarter increase looks dramatic, but the base is much smaller. That is why we would not put a $3.5 million quarterly business in the same growth conversation as a company adding tens of millions of dollars of annual revenue.
| Company | Earlier revenue | Recent revenue | Growth signal |
|---|---|---|---|
| WeRide | $51.6M FY2024 | $97.9M FY2025 | +89.6% |
| Pony.ai | $21.5M Q2 2025 | $36.2M Q2 2026 | +68.8% |
| Trunk Technology | RMB134M FY2023 | RMB345M FY2025 | ~60% two-year CAGR |
| TIER IV | ¥3.87B FY2024 | ¥6.41B FY2025 | +65.6% |
| Autonomous A2Z | ~KRW10.7B 2024 | KRW16B 2025 | ~+50% |
| Kodiak AI | ~$1.8M Q1 2026 | $3.5M Q2 2026 | +91% QoQ |
| Einride | ~$21M H1 2025 equivalent | $27M H1 2026 | +26% |

Funding is rarely spread evenly. See who's soaking up the capital in our autonomous vehicle market report.
Which smaller autonomous vehicle startups are already making money?
Autonomous A2Z, May Mobility and Kodiak are good examples of smaller autonomous vehicle companies that have already moved into real revenue.
Autonomous A2Z reported KRW16 billion of 2025 revenue, with the company saying this was its second consecutive year above KRW10 billion. The business still earns a meaningful amount from government R&D, demonstrations and service projects, so it has not yet reached the same commercial maturity as a scaled robotaxi network.
May Mobility sits at roughly $10 million of annual revenue and has already built a sizable history of commercial rides.
Kodiak is smaller on reported revenue but more advanced in one specific way: customers are already paying for driverless freight operation through customer-owned vehicles.
The lower end of the market has changed. A startup can now generate several million dollars from autonomous-driving products or services without being anywhere close to Waymo's scale.
What separates autonomous vehicle startups making real revenue from the rest?
The autonomous vehicle startups making real revenue have usually narrowed the problem enough that someone can pay for it today.
Waymo concentrates fleets in selected metropolitan areas rather than trying to launch everywhere at once. KargoBot focuses on freight corridors. Gatik works on repeat middle-mile routes between logistics facilities and stores. Kodiak has spent heavily on industrial and freight use cases where paid vehicle hours can build before nationwide driverless trucking exists.
WeRide and Pony.ai take another route by selling several products around autonomous driving while their passenger networks scale.
TIER IV goes further upstream and gets paid for the software, engineering and deployment infrastructure other companies need to build autonomous systems.
There is no single winning business model yet. What the revenue leaders have in common is a customer who can buy something useful now, without waiting for autonomous driving to solve every road and every city.
So who are the real autonomous vehicle revenue leaders right now?
Waymo is currently the strongest candidate for the largest autonomous vehicle business by revenue scale, while WeRide and Pony.ai give us the cleanest evidence of large reported annual revenue.
Waymo's position rests partly on an estimate, so we would not present its revenue as an audited fact. Its operating scale makes the overall conclusion convincing, though: hundreds of thousands of rides each week and millions of annual rides put it well beyond the commercial footprint of most competitors.
WeRide and Pony.ai are the clearest companies to watch if we want hard financial comparisons. Their latest results show businesses approaching or moving beyond the $100 million annualized range while robotaxi revenue itself becomes increasingly meaningful.
Behind them, autonomous freight has produced a real second tier. Trunk Technology, KargoBot and Inceptio have already reached tens of millions of dollars of revenue, while Gatik has accumulated a very large contracted backlog.
Software matters too. TIER IV proves that a company can build a substantial autonomous-driving business without becoming a robotaxi operator.
The market is much more commercial today than the old picture of endless testing suggests. It is also still concentrated, messy to compare and deeply unprofitable. A handful of companies have found serious revenue; far fewer have proved that autonomous driving can become a great business.
OUR METHODOLOGY
This analysis ranks autonomous vehicle startups by the strongest available evidence of current revenue scale. We use reported annual revenue where possible, but also retain clearly labeled run-rate estimates, quarterly revenue and contracted revenue when those figures add useful commercial context.
We keep unlike metrics separate. Waymo's estimated annualized revenue is not treated as equivalent to filed annual revenue, Gatik's contracted revenue is not treated as one year's sales, and operating metrics such as rides, vehicles or loads are not converted into synthetic revenue.
When several figures exist for the same company, metric relevance comes first, followed by source quality and freshness. A recent operating statistic does not automatically replace an older but stronger financial disclosure, while a current quarterly result can be more informative than a stale full-year number for a fast-growing business.
We also check the corporate perimeter behind each datapoint. Autonomous-driving revenue can sit inside a broader automaker, logistics platform, software company or parent group, so companies such as Baidu, Tesla, Mercedes-Benz, BMW and Einride are handled cautiously when a clean standalone autonomy figure is unavailable.
The ranking therefore favors numbers that can reasonably be connected to the autonomous-driving business itself. Filed accounts, SEC disclosures, annual reports, IPO materials and company investor-relations releases carry more weight than third-party estimates, although a credible estimate can still be useful when no standalone company figure exists.
Key sources include Waymo's commercial ride and expansion update, Waymo's operating-scale update, WeRide's FY2025 annual filing, WeRide's Q2 and H1 2026 results, Pony.ai's FY2025 annual filing, and Pony.ai's Q2 and H1 2026 results.
We also used Trunk Technology's HKEX listing materials, Tokyo Stock Exchange listing information for TIER IV, Gatik's financing and contracted-revenue disclosure, May Mobility's SEC transaction materials, Kodiak AI's Q2 2026 SEC disclosure, Autonomous A2Z's company disclosure, and Einride's H1 2026 results.
For companies with large operating footprints but no clean standalone revenue figure, we use those operating disclosures only to explain why they are included or excluded from the numerical leaderboard. Baidu's Q1 2026 and Q2 2026 results are the main example: Apollo Go is clearly large operationally, but we do not invent a revenue figure from ride counts.

Is there a real business behind the hype? Our autonomous vehicle market report digs into the numbers.