Is Tesla losing the robotaxi race today?
In our autonomous vehicle market deck, you will find everything you need to understand the market
SUMMARY
Tesla is not losing the robotaxi race entirely today, but it is clearly behind Waymo and Baidu on the metrics that prove real driverless commercial operations.
The main distinction is between autonomy potential and robotaxi proof. Tesla has the largest supervised FSD data engine, but Waymo has the cleanest paid driverless robotaxi operating curve.
Waymo is currently the most transparent robotaxi company. It discloses rides, miles, weekly activity, city expansion, fleet scale, safety comparisons, and enough methodology to reconstruct its progress.
Tesla likely leads on supervised autonomy miles, with roughly 9B–10B FSD Supervised miles. But those are not fully driverless robotaxi miles, because Tesla itself says FSD Supervised requires active driver supervision.
On fully driverless robotaxi scale, Tesla is far behind. Waymo is near 200M fully autonomous miles, Baidu is around 137M fully driverless miles, while Tesla showed roughly 1.7M cumulative paid Robotaxi miles in Q1 2026.
The best robotaxi KPI is not total miles driven. The cleaner north-star metric is paid driverless rides per week by city, supported by fleet productivity, paid driverless miles, safety events, and remote-assistance rates.
On that KPI, Waymo is the leader. It reports roughly 500,000 paid robotaxi rides per week, while Baidu’s weekly peak is above 350,000 rides and Tesla does not disclose comparable paid driverless ride volume.
Tesla is also not the leader on city footprint, safety proof, or regulation. Baidu appears widest by city count, Waymo has the strongest scaled U.S. commercial footprint, and Waymo has the cleanest rider-only safety case and California regulatory position.
Tesla’s strongest advantage is cost and scaling potential. Its Cybercab target, camera-only strategy, manufacturing base, installed fleet, and OTA distribution could let it compress years of catch-up if the autonomy system clears the driverless threshold.
The catch-up bar is high. Tesla likely needs a 45–50x jump in weekly robotaxi miles, a 50–100x active fleet expansion, 20+ rides per vehicle per day, several cities above 50,000 weekly rides, and Waymo-grade safety disclosure.
So the answer is nuanced: Waymo has the better robotaxi business today, Baidu is the closest scale challenger, and Tesla has the better scaling machine if autonomy finally becomes safely unsupervised.
This market map, featured in our autonomous vehicle market deck, highlights top companies and startups in the autonomous vehicle market
Is Tesla the most transparent robotaxi company right now?
Tesla is not the most transparent robotaxi company today. If we rank companies by usable public metrics, Waymo is #1, with Baidu Apollo Go #2, and Tesla far behind.
We used a simple transparency score: ride volume, autonomous miles, weekly activity, fleet size, city coverage, safety data, growth rate, and methodology. Waymo scores roughly 8/8. Baidu is around 6.5/8. Tesla is closer to 2/8.
Waymo wins because it publishes both scale and pace. Over the last year, it disclosed 100M+, 170M, then nearly 200M fully autonomous miles, plus 400K–500K weekly rides, 4M weekly miles, city expansion, and safety comparisons versus human drivers. That gives us a real operating curve, not just a milestone.
Baidu Apollo Go is the closest challenger. In Q1 2026, it disclosed 3.2M fully driverless rides, a weekly peak above 350K rides, 120%+ YoY ride growth, 22M+ cumulative public rides, 300M autonomous km, and 190M fully driverless km. The numbers are strong, but less externally comparable than Waymo’s, especially on safety methodology and city-level operating domains.
Tesla communicates a lot, but shares little robotaxi operating data. It lists Austin, Dallas, and Houston, and said paid robotaxi miles “nearly doubled sequentially” in Q1 2026.
But without the baseline, that metric is weak. Tesla does not regularly publish robotaxi ride volume, fleet size, total paid miles, driverless miles, remote-assistance rates, intervention rates, or safety outcomes per robotaxi mile.
Tesla may be the loudest robotaxi company, but Waymo is the most transparent robotaxi operator.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Has Tesla driven the most robotaxi miles?
Tesla has probably accumulated the most supervised autonomy miles, but not the most fully driverless robotaxi miles.
That distinction is the whole answer. If we count every mile driven with Tesla FSD Supervised engaged, Tesla is likely far ahead. Tesla’s Q1 2026 chart shows roughly 9B cumulative FSD Supervised miles, and later reporting put the figure around 10B miles. But Tesla itself labels FSD Supervised as requiring active driver supervision. So those are not robotaxi miles in the strict sense.
If we count fully autonomous commercial robotaxi miles, Waymo is ahead. Waymo disclosed nearly 200M fully autonomous miles across 10+ major cities and freeways. That is the cleanest large-scale U.S. robotaxi mileage figure because it refers to fully autonomous operation, not consumer-supervised driving.
Baidu Apollo Go is also ahead of Tesla on driverless robotaxi miles. In Q1 2026, Baidu reported 330M+ autonomous km, including 220M+ fully driverless autonomous km. Converted to miles, that is roughly 205M autonomous miles and 137M fully driverless miles. The disclosure is less directly comparable to Waymo’s, but the order of magnitude is clear.
Tesla’s robotaxi mileage is much smaller. In Q1 2026, Tesla said paid Robotaxi miles nearly doubled sequentially and showed cumulative paid Robotaxi miles rising to roughly 1.7M miles. Even if that is growing fast, it is still far below Waymo’s nearly 200M fully autonomous miles, 117x less, and Baidu’s roughly 137M fully driverless miles, 80x less.
So the clean ranking depends on the definition. On supervised consumer FSD miles, Tesla is #1 by a massive margin. On fully driverless robotaxi miles, Tesla is not #1. Waymo is the clean benchmark, Baidu is comparable in scale, and Tesla is still early.
As this chart shows, and as featured in our autonomous vehicle market deck, search interest in autonomous vehicles has continued to rise
What is actually the best KPI to track robotaxi progress?
Contrary to what most people think, the best robotaxi KPI is not miles driven. It is paid driverless rides per week per city, supported by fleet productivity and intervention-rate data.
That is the cleanest metric because it combines three things at once: technical reliability, regulatory permission, and real customer usage. A company can inflate autonomous miles by driving empty vehicles. It can announce new cities before demand exists. It can show safety charts without proving commercial density.
But if it is doing paid driverless rides every week, in multiple cities, at growing volume, the system is actually becoming a transportation network.
We would track one north-star KPI: weekly paid driverless rides, split by city. “Paid” filters out demos. “Driverless” filters out supervised pilots. “Weekly” shows current operating pace, not a stale cumulative milestone. “By city” prevents companies from hiding behind one strong market while the rest of the rollout is shallow.
The KPI should not stand alone. We would pair it with five anti-gaming metrics: paid driverless miles per week, active robotaxi fleet, rides per vehicle per day, safety-critical events per million miles, and remote-assistance interventions per 1,000 miles.
Together, they reveal whether growth is real, efficient, safe, and increasingly autonomous.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Is Tesla #1 on paid driverless robotaxi rides per week?
No. Waymo is #1 on the clean comparable KPI. Baidu Apollo Go is the closest challenger on scale. Tesla is not rankable because it does not disclose paid driverless ride volume.
Waymo gives the cleanest number: about 500,000 paid robotaxi rides per week across 10 U.S. cities, up from 50,000 weekly paid trips in May 2024.
Waymo also gives a fleet denominator, so we can estimate fleet productivity. With about 3,067 robotaxis in NHTSA filings and an “over 3,000” fleet figure, we can estimate 160+ rides per vehicle per week, or about 23 rides per vehicle per day.
Baidu is close on a weekly scale, but less clean. As we saw, in Q1 2026, it reported 3.2M fully driverless operational rides, a weekly peak above 350,000 rides, 120%+ YoY growth, and 22M+ cumulative public rides. On peak weekly volume, Baidu is at about 70% of Waymo: 350,000 vs 500,000 rides.
On Q1 average, Baidu is further behind: 3.2M rides / ~13 weeks = ~246,000 rides per week, or about 49% of Waymo’s level. We cannot compare rides per vehicle properly because Baidu does not disclose the exact active fully driverless fleet behind those rides.
Tesla does not publish the KPI. In Q1 2026, it said paid Robotaxi miles nearly doubled sequentially and showed cumulative paid robotaxi miles rising to roughly 1.7M miles. But it did not disclose weekly paid rides, city-level rides, active fleet count, rides per vehicle per day, or average trip length.
Waymo first, Baidu second on reported weekly ride scale, Tesla not rankable. Baidu is about 30% behind Waymo on peak weekly rides and about 50% behind on Q1 average weekly rides.
This chart, included in our autonomous vehicle market deck, illustrates yearly VC funding for autonomous vehicle startups
Who has the most usable robotaxi data?
Tesla has the most raw autonomy data. However, Waymo has the most exploitable robotaxi data and Baidu is second.
That is the important distinction. As discussed above, Tesla has accumulated roughly 9B–10B FSD Supervised miles, far more than any robotaxi operator. But those miles come from privately owned cars, with a human still responsible for the drive. They are useful for training autonomy, but they are not pure robotaxi operating data.
For robotaxis, the best data is narrower: paid, driverless, passenger-carrying service. That means real pickups, drop-offs, dense urban routing, curb behavior, fleet redeployment, remote-assistance events, safety incidents, charging, cleaning, and customer usage patterns. This is the data that tells us whether a robotaxi network actually works.
On that definition, Waymo is the leader. As noted earlier, Waymo has disclosed nearly 200M fully autonomous miles, about 4M autonomous miles per week, and roughly 500,000 paid robotaxi rides per week. That is not just driving exposure. It is repeated commercial service at scale.
Baidu Apollo Go is the closest challenger. It has reported roughly 137M fully driverless miles converted from 220M+ fully driverless km, plus 3.2M fully driverless rides in Q1 2026 and a weekly peak above 350,000 rides. The scale is real, but the disclosure is less granular than Waymo’s.
Tesla’s actual robotaxi operating dataset is still much smaller. Its Q1 2026 paid Robotaxi miles were around 1.7M cumulative miles, while its larger FSD dataset remains supervised. So Tesla has the most broad autonomy data, but not the most operational robotaxi data.
So, actually Tesla is still early in true robotaxi operations. Tesla’s advantage is breadth. Waymo’s advantage is purity. For robotaxis, purity matters more than raw volume, because a supervised highway mile does not teach the same thing as a paid driverless pickup in a dense city.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Is Tesla in the most robotaxi cities?
No. Tesla is not present in the most robotaxi cities. Baidu likely has the widest geographic footprint, Waymo has the strongest commercial city footprint, and Tesla is still early with roughly three U.S. markets.
Tesla’s robotaxi footprint is actually still small. It has activity in Austin, with expansion mentioned in Dallas and Houston. So even if we count all three, Tesla is at roughly 3 U.S. cities.
That is far behind the leaders. Waymo says it operates across about 10 U.S. commercial metro areas, including Phoenix, San Francisco, Los Angeles, Austin, Atlanta, Miami, Dallas, Houston, San Antonio, and Orlando. Baidu Apollo Go has an even wider footprint if we count broad robotaxi operations, with activity reported across roughly 20 Chinese cities, plus expansion in the Middle East.
The nuance is that city count can be misleading. A company can announce a city with a small geofence, limited vehicles, or limited hours. That is why we separate announced presence from scaled commercial service.
But, on both versions, Tesla is not first: Baidu appears widest by city count, while Waymo has the cleanest scaled commercial footprint.
This chart, included in our autonomous vehicle market deck, shows how Waymo is winning in autonomous vehicles
Is Tesla the safest robotaxi company?
No. Tesla has the largest supervised-safety dataset but Waymo has the strongest proven robotaxi safety case.
The key distinction is the same as before: Tesla reports safety for FSD Supervised, while Waymo reports safety for rider-only driverless service. Those are not the same category. FSD Supervised still has a human responsible for monitoring the drive. A robotaxi safety claim should be based on vehicles operating without a human driver.
Tesla’s safety data is strong, but it is not pure robotaxi data. Tesla reports more than 10.9B FSD Supervised miles, including about 4.1B city miles, and publishes collision rates for FSD Supervised versus manually driven Teslas and U.S. averages. That is useful evidence that FSD Supervised may reduce collision frequency.
But Tesla does not yet publish the full robotaxi safety table we would need: crashes per million paid driverless miles, injury crashes, disengagements, remote-assistance events, and safety-critical incidents per robotaxi mile.
Waymo’s case is cleaner. Through December 2025, Waymo reported 170.7M rider-only miles with no human driver. It claims 92% fewer serious-injury-or-worse crashes, 83% fewer airbag-deployment crashes, and 82% fewer injury-causing crashes versus human benchmarks in its operating areas.
It also reports vulnerable-road-user reductions: 92% fewer pedestrian injury crashes, 85% fewer cyclist injury crashes, and 81% fewer motorcycle injury crashes.
Baidu Apollo Go has a large driverless scale, but less comparable safety disclosure. It reported 220M+ fully driverless autonomous km, around 137M miles, and called its safety record “outstanding.”
But it does not provide the same public crash-rate breakdown as Waymo: serious injuries per million miles, airbag deployments, injury crashes, or city-adjusted human benchmarks.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Is Tesla the most advanced robotaxi company on regulation?
No. Tesla is not the most advanced robotaxi company on regulation. In the U.S., Waymo is clearly ahead.
Global comparisons are messy because China, the U.S., and the Middle East use different approval systems. So the cleanest comparison is California, where robotaxi companies need both DMV autonomous approval and CPUC passenger-service approval.
Waymo has cleared both layers in California: DMV deployment authorization and CPUC driverless deployment authority for paid passenger service. That is the strongest U.S. regulatory position.
Tesla has not. In California, Tesla Robotaxi LLC is listed for testing with a driver, but not for driverless testing, not for deployment, and not for CPUC driverless deployment. So Tesla cannot claim the same regulatory status as Waymo in the strictest U.S. market.
Tesla is further ahead in Texas. Tesla Robotaxi LLC received a Texas transportation network company license valid until August 6, 2026, and has been rolling out in Austin, Dallas, and Houston. But Texas is a lighter regime than California.
Other players matter, but do not change the ranking. Zoox and WeRide have California driverless pilot permits. Nuro has approvals. Cruise used to be relevant, but its California driverless permits were suspended after its 2023 incident and GM later wound down the robotaxi program.
So, Tesla has regulatory momentum in permissive markets. Waymo has deeper regulatory proof in stricter markets.
This chart, included in our autonomous vehicle market deck, illustrates yearly funding for autonomous vehicle startups
Who is accelerating fastest in robotaxis right now?
Tesla is probably accelerating fastest from a tiny base, but Waymo is accelerating fastest at real scale.
Tesla said paid Robotaxi miles nearly doubled sequentially in Q1 2026, and cumulative paid Robotaxi miles rose from roughly 610,000 miles at the end of Q4 2025 to about 1.7M miles at the end of Q1 2026. That is nearly 3x cumulative growth in one quarter. On pure growth rate, Tesla looks very fast.
But the base is still small. As discussed above, Tesla does not disclose weekly rides, active fleet count, or city-level volume. So the growth is real, but hard to benchmark against Waymo or Baidu.
Waymo is the stronger acceleration story at scale. Its paid weekly rides grew from about 50,000 per week in May 2024 to around 500,000 per week in early 2026. That is roughly 10x growth in less than two years, while expanding from its original core markets into 10 U.S. rider markets. It is not just growing quickly, but also from an already large base.
Baidu Apollo Go is also accelerating hard. In Q1 2026, it reported 3.2M fully driverless rides, weekly rides peaking above 350,000, and 120%+ YoY ride growth. That makes Baidu the closest non-U.S. scale comparison, but its city-level and fleet productivity data is less clean than Waymo’s.
In conclusion, Tesla has the fastest percentage growth if we only look at paid Robotaxi miles, but Waymo has the best acceleration at commercial scale.
If you want more recent data on this point, please see our latest autonomous vehicle market report.
Who will make robotaxis the cheapest in the long run?
Tesla has the best shot at the lowest long-term cost per mile, but Baidu has the strongest proven low-cost robotaxi vehicle today.
The cost comparison is not about ride price today. It is about unit cost and cost per mile: vehicle capex, autonomy hardware, sensors, energy, maintenance, insurance, cleaning, remote assistance, and utilization.
Tesla’s cost case is the most aggressive. The Cybercab is designed as a purpose-built two-seat robotaxi, with no steering wheel or pedals, a target price below $30,000, and a claimed long-term operating cost of about $0.20 per mile, or $0.30–$0.40 per mile including taxes and other costs. If Tesla gets there, it would likely be the lowest-cost robotaxi platform.
Baidu is the best current low-cost benchmark. Its Apollo RT6 production cost has been reported at 250,000 yuan, or roughly $34,500, with earlier launch figures around 204,600 yuan, or about $28,350. So Tesla’s sub-$30,000 Cybercab target is roughly 13% cheaper than Baidu’s $34,500 figure, but roughly in line with Baidu’s earlier $28,350 figure.
Waymo is harder to price because it does not disclose full vehicle-plus-autonomy cost. But the direction is clear: it is moving down from an expensive base. The old Jaguar I-PACE started around $72,000 before Waymo’s autonomous hardware. Waymo’s new Ojai, built with Zeekr, reportedly cuts sensors by 42% and costs about $75,000 less than the Jaguar-based platform. The new stack still uses 13 cameras, 6 radars, and 4 lidars, so it remains richer than Tesla’s camera-only approach.
That puts the comparison in ranges. Against Baidu, Tesla’s target is only 0–15% cheaper on vehicle capex, depending which Apollo RT6 cost figure we use.
Against Waymo’s old Jaguar-based platform, Tesla could be well over 50% cheaper on vehicle cost before even counting sensor hardware. Against Waymo’s new Ojai, the gap is smaller but still likely meaningful, because Tesla is targeting sub-$30,000 while Waymo still carries lidar, radar, integration, and a larger vehicle format.
This chart, included in our autonomous vehicle market deck, compares the main business model options for autonomous trucking companies
Where is Tesla still behind others in robotaxis?
Tesla is still behind the leading robotaxi players on the metrics that prove real operations: rides, driverless miles, city footprint, safety evidence, regulation, and proven low-cost hardware.
The gap is not the same everywhere.
Against Waymo, Tesla is mostly behind on operational maturity. Against Baidu, Tesla is behind on proven low-cost robotaxi hardware and broad geographic deployment.
In the table below, we only include areas where Tesla is actually behind.
| Category | Estimated delay vs leader | Why this estimate |
|---|---|---|
| Paid driverless ride volume | 2–4 years behind Waymo | Waymo is at roughly 500,000 paid rides/week, up from 50,000/week in May 2024. Tesla does not disclose weekly paid rides, so it has not reached a comparable public baseline yet. |
| Fully driverless robotaxi miles | 3–4 years behind Waymo | Waymo is near 200M fully autonomous miles. Tesla showed roughly 1.7M cumulative paid Robotaxi miles in Q1 2026, about 117x less. |
| Operational robotaxi data | 3–4 years behind Waymo | Waymo has large-scale data from paid driverless pickups, drop-offs, dense city routing, fleet redeployment, and passenger service. Tesla has far more supervised FSD miles, but much less pure robotaxi operating data. |
| Weekly ride scale | 2–3 years behind Waymo | Waymo is at 500,000 paid rides/week. Baidu peaks above 350,000/week. Tesla does not publish weekly ride count, so it cannot be ranked on the main usage metric. |
| Commercial city footprint | 1–3 years behind Baidu / Waymo | Tesla is around Austin, Dallas, and Houston. Waymo is around 10 U.S. commercial metro areas. Baidu has activity across roughly 20 Chinese cities, depending how broadly we count robotaxi operations. |
| Safety proof | 3+ years behind Waymo | Waymo reports safety outcomes across 170M+ rider-only miles, including injury and airbag-deployment crash reductions versus human benchmarks. Tesla reports FSD Supervised safety, but not a full paid driverless robotaxi safety table. |
| California regulation | 2–3+ years behind Waymo | Waymo has California DMV deployment and CPUC driverless deployment authority. Tesla Robotaxi LLC is listed for testing with a driver, not driverless deployment or paid driverless passenger service. |
| Transparency | 2–3 years behind Waymo | Waymo publishes rides, miles, weekly activity, fleet scale, safety methodology, and city expansion. Tesla mostly publishes paid Robotaxi miles and high-level rollout language. |
| Proven low-cost robotaxi hardware | 1–2 years behind Baidu | Baidu’s Apollo RT6 has reported production-cost figures around $28,000–$34,500. Tesla’s Cybercab target is sub-$30,000, but it is not yet proven in volume production or scaled robotaxi service. |
What would Tesla need to catch up quickly in robotaxis?
Tesla can catch up quickly only if it compresses years of robotaxi operations into the next 12–18 months.
As shown above, Tesla is not missing one thing but operating density. The gap is now measurable: weekly rides, weekly miles, active robotaxis, city depth, safety disclosure, and regulatory depth.
The first target is miles. Tesla was around 1.7M cumulative paid Robotaxi miles in Q1 2026, while Waymo is near 4M fully autonomous miles per week. So Tesla would need to reach roughly 2–3 Waymo-style weeks of output every week just to stop falling behind, and a 45–50x increase in weekly robotaxi miles to match Waymo’s current operating pace.
The second target is fleet size. Public reporting put Tesla at dozens of registered robotaxis in Texas, while Waymo is around 3,000+ vehicles. So the catch-up requirement is not “add more cars.” It is more like 50–100x active fleet expansion, while keeping the service driverless and reliable.
The third target is utilization. Waymo’s rough productivity is about 23 rides per vehicle per day. That gives Tesla a concrete target: not just 1,000 robotaxis, but 1,000 robotaxis doing 20,000+ paid rides per day. If Tesla reaches 3,000 vehicles at that utilization, it could be near 500,000 weekly rides, which is today’s Waymo scale.
The fourth target is city density. Tesla being in Austin, Dallas, and Houston is not enough if those markets stay thin. To close the gap, Tesla needs at least 5–10 cities with real ride volume, not just service-area announcements. A useful benchmark would be 50,000+ paid rides per week in several cities, because that was Waymo’s May 2024 level before it scaled to 500,000.
The fifth target is safety and regulation. Tesla cannot catch up reputationally with mileage alone. It needs a public robotaxi safety table: crashes per million driverless miles, injury crashes, remote-assistance events, and safety-critical interventions. It also needs at least one hard-market approval comparable to California driverless deployment, not only faster expansion in permissive states.
So, to catch up with others, Tesla needs a 45–50x jump in weekly robotaxi miles, a 50–100x fleet expansion, 20+ rides per vehicle per day, several cities above 50,000 weekly rides, and Waymo-grade safety disclosure.
This chart, featured in our autonomous vehicle market deck, shows the share of revenue generated by each customer segment in the autonomous vehicle market
Where is Tesla still ahead in robotaxis?
Tesla is not the robotaxi leader today, but it is ahead on the inputs that could make it catch up unusually fast.
The gap is simple.
Waymo leads in proven robotaxi operations, but Tesla leads in scalable ingredients: supervised driving data, vehicle manufacturing, installed fleet, hardware simplicity, and long-term cost potential.
Those advantages are not enough to declare Tesla the leader. But they explain why Tesla could compress a 3–4 year operating gap faster than a normal AV company.
| Category | Estimated lead vs closest rival | Why Tesla is ahead |
|---|---|---|
| Raw autonomy data | 5–10 years ahead of Waymo | Tesla has roughly 10.9B FSD Supervised miles, including 4.1B city miles. Waymo has around 200M fully autonomous miles. Tesla’s data is less pure, but about 50x larger in raw mileage. |
| Consumer AV fleet | 5+ years ahead of Waymo / Baidu outside China | Tesla has millions of vehicles already on the road and more than 1M FSD users / FSD-capable active vehicles in recent reporting. Waymo’s fleet is around 3,000+ robotaxis. Tesla’s fleet is not driverless, but it is already deployed hardware at mass scale. |
| Manufacturing scale | 3–5 years ahead of Waymo | Tesla produced 434,358 vehicles in Q4 2025 alone. Waymo still depends on partners such as Jaguar and Zeekr for robotaxi platforms. If Tesla converts even a small share of its production into robotaxis, the scaling ceiling is much higher. |
| Hardware cost philosophy | 2–4 years ahead of Waymo | Tesla is building around a camera-only approach. Waymo’s newer stack has improved, but still uses 13 cameras, 6 radars, and 4 lidars. Tesla’s bet is riskier, but structurally cheaper if it works. |
| Long-term unit cost | 1–3 years ahead of Waymo, roughly tied with Baidu target range | Tesla’s Cybercab target is sub-$30,000 and $0.20/mile operating cost, or $0.30–$0.40/mile all-in. Baidu’s Apollo RT6 is already reported around $28,000–$34,500. Waymo’s platform is likely more expensive because of vehicle format, sensor stack, and U.S. operating costs. |
| OTA distribution | 5+ years ahead of traditional robotaxi operators | Tesla can push software to a large consumer fleet. Waymo improves a dedicated fleet; Tesla can update millions of cars. That does not make them robotaxis, but it makes iteration faster if the same autonomy stack transfers. |
| Speed from small base | Potentially ahead, but not proven | Tesla’s paid Robotaxi miles grew from roughly 610,000 to 1.7M in one quarter, nearly 3x cumulative growth. Waymo’s growth is stronger at scale, but Tesla’s early slope is steep. The issue is whether this compounds beyond a small base. |
Tesla is ahead on scale potential, not robotaxi proof. It has the largest data engine, the largest vehicle-manufacturing base, the simplest hardware-cost strategy, and the strongest theoretical cost curve.
But we should not confuse these advantages with operational leadership. Tesla’s raw FSD miles are supervised. Its fleet is mostly privately owned. Its Cybercab cost is still a target. Its OTA advantage only matters if the system becomes safely unsupervised.
Tesla’s lead is not in today’s robotaxi network. It is in the possibility of turning a car company into a robotaxi deployment machine. Waymo has the better robotaxi business today. Tesla has the better scaling machine if autonomy finally clears the threshold.
OUR METHODOLOGY
This analysis tests whether Tesla is losing the robotaxi race today based on disclosed driverless commercial operation, not broader autonomy activity. We separated supervised FSD miles, paid robotaxi miles, rider-only miles, and driverless ride volume because they describe different levels of autonomy, operating risk, and commercial proof.
For transparency, we looked at the public metrics that let a reader reconstruct the operating curve: rides, miles, fleet scale, city footprint, safety outcomes, growth rate, and methodology. We gave more weight to disclosures that connect several of those signals, rather than isolated milestones.
We included Baidu Apollo Go as a scale peer because its reported ride volume and driverless-mile base are large enough to matter. We kept it separate from Waymo where the disclosure is less granular, especially around safety methodology, city-level operating domains, and fleet productivity.
For city footprint, we separated announced presence from scaled commercial service. A city only counts as strategically meaningful when it shows real operating density, not just a small pilot, a narrow geofence, or a rollout announcement.
For safety, we compared evidence by operating mode. Tesla’s FSD Supervised data is useful for understanding supervised autonomy performance, while Waymo’s rider-only data is the cleaner robotaxi safety benchmark because the vehicle is operating without a human driver.
The delay and lead estimates are directional operating-gap ranges, not company-reported figures. We based them on the size of the current gap, reported growth rates, public deployment maturity, and whether the underlying metric is already proven at commercial scale.
For cost, we separated proven vehicle-cost benchmarks from company targets. Baidu’s RT6 is treated as the strongest current low-cost robotaxi benchmark, while Tesla’s Cybercab is treated as a long-term cost target until it is produced and operated at scale. Waymo’s cost position is interpreted directionally from its vehicle format and sensor stack, rather than from a disclosed full platform cost.
Key sources used for this analysis include: Waymo safety impact data and methodology, Waymo’s 2024 disclosure of 50,000+ paid weekly rides, Waymo’s 2025 disclosure of 250,000+ paid weekly trips and fleet scaling, Waymo’s 2026 market expansion and one-million-rides target, Waymo’s sixth-generation sensor stack, Baidu Q1 2026 results, including Apollo Go ride volume, Baidu Apollo RT6 production-cost disclosure, Tesla Q1 2026 update, including paid Robotaxi miles, Tesla FSD Supervised safety report and mileage counter, California DMV autonomous vehicle program and permit categories, California DMV autonomous vehicle permit holders, and CPUC driverless autonomous vehicle deployment program status.
This chart, included in our autonomous vehicle market deck, shows how robotaxi platform technology has evolved over time
Related blog posts
- How strong is fundraising in the autonomous vehicle market right now?
- The startups that have raised the most funding in the autonomous vehicle market
- The most highly valued startups in the autonomous vehicle market
