How the robotaxi cost calculator works
Here is the thing almost every Cybercab argument gets wrong, on both sides: a robotaxi does not compete against "owning a car" as one number. It competes against two very different numbers, and which one applies depends entirely on a decision you have to make first. If you would keep your car anyway, a robotaxi ride competes against your car's marginal cost, just the fuel, tires, and wear, about 24 cents per mile on AAA's 2025 averages, because the insurance, depreciation, and registration are already spent whether the car moves or not. Almost nothing beats 24 cents. But if a robotaxi would let you give up a car entirely, or skip buying a second one, it competes against the full cost of ownership, fixed costs included, which runs about 77 cents per mile at average American mileage and climbs fast as the miles fall. That is a fight a robotaxi can genuinely win, and this calculator runs both fights side by side so you can see which one you are actually in.
Enter the miles a robotaxi would cover, your car's annual miles, and a fare: a preset from the real world (labeled honestly as measured fares or company targets, because those are not the same thing) or your own number. Use AAA's averages for the car, or switch to detail mode and enter your own annual and per-mile figures. You get both scenario ledgers and the number worth sharing: the break-even fare below which dropping the car wins on money.
The formula
full annual cost = fixed costs (insurance + registration + depreciation + finance) + marginal × annual miles
scenario A (keep the car): robotaxi miles × fare vs robotaxi miles × marginal
scenario B (drop the car): annual miles × fare vs full annual cost
break-even fare = full annual cost ÷ annual miles
The defaults are AAA's 2025 Your Driving Costs figures for a new vehicle: fuel 13.00 cents per mile, maintenance, repairs, and tires 11.04 cents, insurance $1,694 a year, license, registration, and taxes $813, depreciation $4,334, and finance charges $1,131. Those components sum to within a dollar of AAA's published $11,577 annual total at the study's 15,000 miles a year, about 77 cents per mile. Every one of them is editable in detail mode, because your car is not an average.
Worked example
A robotaxi at a flat $1.00 per mile, covering 4,000 miles of trips, against a 12,000 mile car on AAA average costs.
Scenario A, keep the car: 4,000 miles × $1.00 = $4,000 a year, against 4,000 × 24.04 cents = $961.60 of marginal driving cost. The robotaxi costs $3,038.40 more per year. At this price you are buying back the driving time, not saving money.
Scenario B, drop the car: the full cost of the car is $7,972 of fixed costs plus 12,000 × 24.04 cents = $2,884.80 of driving, so $10,856.80 a year. Robotaxi for all 12,000 miles: $12,000. Owning still wins, but only by $1,143.20 a year. The break-even fare is $10,856.80 ÷ 12,000 = 90.5 cents per mile: below that, dropping the car wins on money.
Now run Tesla's stated Cybercab target of 35 cents per mile through the same machine. Scenario A: $1,400 against $961.60, so the robotaxi still loses by $438.40 a year against the car you keep, even at Tesla's own aspirational price. Scenario B: $4,200 against $10,856.80, so dropping the car saves $6,656.80 a year. Same fare, opposite verdicts. That is the entire robotaxi economics debate in one line: the target price loses to the car in your driveway and demolishes the car you could sell.
Cybercab: what has been claimed, and what is actually charged
This page's credibility hangs on one distinction, so here it is plainly: a target is a hope with a press event, and a meter is a price. At the We, Robot event in October 2024, Elon Musk said the Cybercab's operating cost would come in around 20 cents per mile over time, that rides might be priced around 30 to 40 cents per mile, that the vehicle would sell for under $30,000, and that production would start before 2027. Those are company targets, stated on stage, and this calculator labels them that way every time you pick one.
What the meters actually say as of August 2026: Tesla's Austin robotaxi service charges a $3.00 base plus $1.40 per mile (March 2026, after two price increases in quick succession from its original $1 base plus $1 per mile), which works out to about $2.00 per mile effective on a 5 mile trip. Waymo, the most established robotaxi operator, prices like a slightly premium rideshare: spring 2026 app pulls put the average Waymo trip at $19.69 against $17.47 for a comparable UberX, with San Francisco's rate card near a $4.95 base plus $3.65 per mile; the premium over Uber shrank from 30 to 40 percent in mid-2025 to about 13 percent, and nobody tips a robot. Ordinary rideshare itself lands around $2 to $3 per mile all-in for typical urban trips: the rate card's $1 to $2 per mile looks cheaper than reality precisely because it omits the base fare, time charges, booking fees, and tip. So the real robotaxi market today prices at five to ten times Tesla's stated target. That does not make the target impossible. It makes it a target.
Why the car you already own is so hard to beat, and why the second car is the real target
The reason is in AAA's own breakdown: ownership is mostly fixed. Insurance, depreciation, and finance charges alone are $7,159 of the $11,577 average year, about 62 percent, and adding registration brings the fixed share to roughly 69 percent. All of it is spent before the car moves an inch. So when you already own the car, a robotaxi ride is competing with 24 cents a mile, and even Tesla's 35 cent aspiration loses that fight. This is why "robotaxis will replace car ownership" arguments that compare fares against the full 77 cents quietly assume the conclusion: the full number only applies if the car actually goes away.
Which is exactly why the second car is the sharpest version of the question, for two reasons that compound. Second cars drive fewer miles, and the fewer miles a car drives, the worse its full cost per mile gets, because the fixed costs are spread over less driving: a car doing 5,000 miles a year on AAA averages carries a break-even near $1.83 per mile, a number today's measured fares are already brushing against. A lightly used second car is the most expensive per-mile vehicle in America, and it is the one a robotaxi can plausibly replace this decade.
And the utilization argument deserves its honest due, because it is why fleet robotaxis could eventually get under private ownership costs: a private car sits parked about 95 percent of the time (Paul Barter checked the folklore three separate ways and got 95 percent or more each time; the RAC Foundation finds the average car moving about one hour in 24), while its fixed costs run around the clock. A fleet car spreads insurance and depreciation across many paying trips a day. The honest unknowns sit between that theory and a price on a meter: what insurance costs for a driverless fleet, what cleaning and vandalism cost when nobody is watching the back seat, what remote-operator support costs, and what surge pricing does to the fare at exactly the hours you need a car. For a sanity check from a different direction: the IRS reimburses business driving at 76 cents per mile in the second half of 2026, a full-cost proxy within a couple of cents of AAA's figure. Our mileage reimbursement calculator keeps that number current.
What the dollars cannot price
One thing this page refuses to do is pretend money is the whole answer. An hour in the back seat reading, working, or dozing is worth something, and for a long commute it may be worth more than the entire cost difference. Waiting eight minutes on a curb for a car that may surge-price you is worth something negative. A car in the driveway at 2 a.m. when a kid is sick has a value no fare table captures, and so does never having to find parking downtown again. This calculator keeps the ledger to dollars, states so in the result, and leaves the time argument where it belongs: with you, weighed on top of an honest number instead of hidden inside a dishonest one.