Tesla Model 3 Teardown Reveals It’s Relatively Cheap To Produce


With CEO Musk’s recent Tweets about Tesla Model 3 production costs and release of certain variants, this new information is even more compelling.

We hear time and time again that automakers are losing money on EVs. Some people believe that an actual dollar amount is lost or absorbed every time an EV is sold. It’s not that simple. There are so many variables involved here when factoring in development costs, volume, etc. It’s a huge, tricky math problem that’s not so easy to comprehend.

With this being said, Tesla CEO Elon Musk’s recent Tweets about holding off on lower-priced models until volume is higher – so Tesla doesn’t die – makes perfect sense. For a myriad of reasons, as production increases, the automaker should be able to more easily turn a reasonable profit on less expensive trims. Now, a German firm that specializes in vehicle teardowns has put a number out there.

The German company recently spoke with WirtschaftsWoche. An engineer shared:

If Tesla manages to build the planned 10,000 pieces a week, the Model 3 will deliver a significant positive contribution to earnings.

In fact, it was revealed that the materials and logistics involved in building the Model 3 add up to about $18,000. Labor costs were determined to be about $10,000. This may not include some of the other factors that we spoke to above, but it’s pretty telling, nonetheless. It correlates well with Musk’s recent Tweet confirming that Model 3 production cost could definitely come down to ~$28,000 once production reaches 10,000 units per week.

The teardown expert also made references to battery cells. The 2170 cells used in the Model 3 battery pack are made up of only 2.8 percent Cobalt. The industry average is 8 percent, which means the Model 3 battery pack requires 65 less percent Cobalt than the average. Being that the material is hard to come by, this may give Tesla another advantage. Managing Director of Batterien-Montage-Zentrum (BMZ) Sven Bauer said:

That would be a significant competitive advantage for Tesla. Cobalt is currently very difficult to get on the world market.

Source: Teslarati, Electrek

Categories: Tesla

Tags: , ,

Leave a Reply

83 Comments on "Tesla Model 3 Teardown Reveals It’s Relatively Cheap To Produce"

newest oldest most voted

I hear they saved a bunch on the 12 volt charging system. Sorry, I couldn’t resist. 🙂

Cruel, but true.

It was the copper wire saved by redesigning 12V circuit

One by one the shorts arguments are disappearing:)

LOL …I think they all Bailed and Lost their asses…Shorting is gambling and Should be Outlawed Because it Cre

“Cheaper to make than most think”

When I look at the Model 3 I think it looks really cheap to make. Look at the interior design. Flat simple dash with no buttons, symmetrical left to right. Nothing looks really difficult to form.

Electric drivetrains are pretty simple, and it sounds like they have made it simpler to make. Power comes for free with larger batteries.

The battery is the expensive part.

Also, Musk stated that they significantly improved the cabling (communication, high-voltage, low-voltage) architecture so that the Mod3 requires only a fraction of the cabling the ModS requires.

Yes, the cutbacks in electric are obvious… No heated seat controls for rear seat passengers, they have to ask the driver to make adjustments for them in the touch screen… Thats not a good solution…

Or they can just implement voice control for the same thing through an OTA “Dave”.

You like so many, are constantly constraining yourself with you lack of thinking outside the box.

haha! So voice control for the rear seat passengers… tell me how that works.. “Be quiet everyone..”. “Hey Siri, I am in the Left rear seat, can you increase the seat heating 1 notch”

I’d rather just reach up over and set it myself… quietly, and without having to ask everyone in the car to be quiet… I-Pace you can control your heat, fan, and seat warming function from the back seat. Or if is is kids, parents can disarm the rear seat controls. .

This was just done for cost cutting , and it sucks… IMO

I see, Dave. So Tesla cops all sorts of flak for only building toys for rich boys but, when they develop a cheaper alternative, you complain that they’ve cut back on luxury features.

You should thank Tesla for forcing Jaguar (and all the rest) to finally come up with some competition which you can benefit from.

Well, every EV on the market that I have been in has heated rear seats with controls in the back, this is a very strange miss with the Tesla, because its the first car to not have controls in the back for the rear passengers… Corners cut the the wrong place.

This is the level of decadence we have reached in 2018: dismissing a car because the rear seat passengers do not have a switch for bum warmers.

I dismiss Model 3 for many reasons, I drove an early one and the quality was not terrific, I do not like the interface… Pretty much read consumer reports review, and that is where I am at. Back seat comfort should be better, and controls should be better . I am not the only one feeling that way, most reviewers have repeated the same comments I have over and over again.

Well, luckily we live in a world where opinions differ. For Tesla it’s of no consequence that ‘Dave’ dislikes the Model 3. What counts is how many people like it. That’s all Tesla needs.

That’s all any company needs.

That is a super arrogant statement. With Tesla there is definitely a true “reality distortion field” and many people ordering the model 3, have never driven one, or even sat in one, so they do not know how it will work in daily life. Tesla has done a great job of hyping all of their products (which are good, but not great), and the real car manufacturers have been slow to come forward with compelling EV’s. This is in transition, as the larger makers come forward with well made BEV’s Tesla will have to compete on real merits, and not just hype. Like I said, this is in transition now, and on the luxury side I think Tesla is about to be assaulted from several directions. Arrogant statements like yours are exactly what is going to get Tesla in trouble long term. Tesla better be closely watching the rearview mirror as the steamroller is coming from behind and gaining steam…

I couldn’t care less. None of my previous cars even had rear heated seats, and the amount of time there is even anyone in the back seat is so infrequent this really isn’t a problem for me.
Fair enough some people have a problem with it, but I think on balance it will be pretty small problem.

haha! So voice control for the rear seat passengers… tell me how that works.. “Be quiet everyone..”. “Hey Siri, I am in the Left rear seat, can you increase the seat heating 1 notch”

I’d rather just reach up over and set it myself… quietly, and without having to ask everyone in the car to be quiet… I-Pace you can control your heat, fan, and seat warming function from the back seat. Or if is is kids, parents can disarm the rear seat controls. .

This was just done for cost cutting , and it sucks… IMO

You highlight a problem with voice controls in general, not just from those in the back seat. At least one report from an actual user said he had to turn the car’s stereo down, and tell the kids in the back to shut up, before he could get the voice control to react properly. I presume that’s a general problem with the tech, not just an anecdotal report.

I think the industry is going to move towards gesture control, because voice control is too unreliable. Gesture control could be implemented in the back seat, too, with a ceiling-mounted camera above the back seat.

In my experience (Model S) the stereo is muted when voice control is activated. Misbehaving children in the back seat is surely a parental discipline issue – perhaps not a shortcoming on Tesla’s part. 🙂

I should have added the voice control user report I remember wasn’t from a Tesla owner; I think it was a Mercedes owner. But yeah, noisy kids in the back seat is not actually a shortcoming on the part of the auto maker. 🙂

Okay, so the Model S mutes the stereo when voice control is activated. But isn’t that a chicken-and-the-egg problem? How well does the car “hear” you trying to activate the voice control if the music is turned up loud?

Or is there a specific button/knob press on the steering wheel buttons/knobs that tells the car to activate voice control?

How about just some damn buttons, knobs and switches??!! They have served us well for 100 years. I know, I know, too “inside the box”, too old school. So 20th century… blah, blah, blah. If Tesla could offer us a “Crusty old geezer” edition, I might be interested.

Me too, and I think the control interface as it is has real safety issues, too much eyes off the road. Don’t make things take two or more steps that could be done by one. Don’t make it necessary to look when it can be done by feel.

Enjoy your iPace!

I think I-Pace found the right mix of touch screens and knobs…

The I-Pace is a very nice daily driver but at two-thirds the range of the Model 3, it’s going to driver you crazy on road trips.

That may be true… I do not take many road trips, but realistically I do take trips out of the I-Pace range, so I will find out about the charging issues, and certainly report back my experience.

“I think the industry is going to move towards gesture control, because voice control is too unreliable”

You act as if voice control is what it is. No. The tech is constantly improving and we have only scratched the surface of what is possible. We are in in the Commodore-64 age of conversational interfaces.

For instance, by using more than one microphone, it is astounding how the voice of a single person among a crowd can be isolated.

AS far as I am concerned, Voice control is a dead end because the user needs to know that the command exists in advance (from reading the manual) before being able to use it.
A button on the dash teaches the user that the feature exists over and over again just by doing something else and looking at the dash, and using the feature is simple and self explanatory.

There is a similar issue with computers between command line interfaces and graphical user interfaces.
Command lines are faster, more powerful, and more efficient, but there is a very steep learning curve, you need to learn the commands and the syntax in advance.
Graphical interfaces are more user friendly, and if designed properly, teaches the users how to use the computer without them realizing it. But it takes a lot of time to design one properly, and it takes more time to use the feature, browsing through menus, tabs, buttons, etc…

Yeah, those buttons with icons that you have to look up in the manual to see what they mean, or you have press and find out what happens. Yes, very intuitive indeed. 😉

Or how about an app with a guest pass for rear seat passengers that allows them limited access to only those functions needed by those in the rear.

I had a Nissan LEAF for 6 years (April 2012 – April 2018). I can count on 1 hand the number of times the rear seat heater was used. The lack of local control for rear seat heat is a non-issue.

I can’t beat you (March 2013 purchase) but I can confirm – 1 hand. It was actually 1 finger.
In my world, most back seat passengers are in seats that negate the heating. But at no point in my life was it common to have adult back seat passengers driving around in the cold. But I did have a 2 seater when I was younger….

They shouldn’t, leave the controls to the driver, for safety reasons. If the car ever becomes autonomous then you could do that.

Tesla’s are not going to be autonomous anytime in the near future. That is 5 years to a decade away, first they have to stop running into emergency vehicles, and barricades, high speed automatic emergency braking would be a nice start.

If you stopped carpet bombing all Tesla-related stories at InsideEVs with FUD, that would be a nice start. If you’re losing money on your TSLA short investment, that’s your problem… not ours.

Just the truth, Tesla is way behind in this Tech, and now they really have no money to do deep investment. Cruise Automation has more money committed in the last week, then Tesla has in their entire cash pile… Just the facts guys, its not picking on Tesla.

And the reason why is the time its going to take for the legislation to catch up with the technology – not because manufactures like Tesla cant design and build an autonomous vehicle but because changes in law and deciding who is responsible in a fatal autonomous vehicle accident. This is something many law experts have commented on.

Disagree on this one… Leaving that out, was just cheap, and not a good option.

Definitely a “first world problem”. I’ve only been in one single car which had separate climate controls in the rear, and that was a minivan. It’s no big deal to ask someone in the front seat to adjust the controls when needed.

Good for Tesla in saving money by not including that completely unnecessary luxury in this sedan.

Its the only car on the market that has this problem… and its a big one, people who own Model 3 complain about this…

Might get added to the app. Certainly not a show-stopper, but some need to whine about something.

Since 99% of the cars on the road don’t even have a REAR seat heating system I think we will be ok.

Actually I cannot name any EV that does not have rear seat heating… Bolt, Volt, and Leaf all have it

Chevy Spark EV does not have rear seat heating. And controls for the front seats are either on or off, switched by a button that you have to bend over and stretch to reach.

That’s because you are new to this and don’t know the evs on tne market.

Sorry Mark, in 2018, they all have heated rear seats… And on the entire auto business there is not one car with rear heated seats that does not have controls in the back…

Bottom line is this is a terrible mistake, Tesla being cheap…

Must be market dependant. My Leaf doesn’t have heated seats at all, let alone in the rear. Used to have heated seats in my convertible, thought I’d never NOT want them again, but you know when the car is pre heated before you even hey in, heated seats just aren’t that much of an issue.
The Leaf made a big deal about heated seats, and that was because the range is so low you want to save as much power for driving as possible. Somehow, I don’t think that will be as much of a problem in a Tesla.

Isn’t a fix already in the works for this? I’ve heard mention of this being controlled by an app (and that multiple phones can connect to the car at the same time).

So a rear seat passenger needs to log into the car to turn their seat warmer on? Really? That sounds simple… I think the voice control would be more simple

Buttons are actually not very expensive, but I agree on the rest. Especially compared to the S and X(!!) it’s a pretty simple car to build.

It’s not the button. It’s the wiring to the button that’s expensive.

“…it was revealed that the materials and logistics involved in building the Model 3 add up to about $18,000…”
Fun fact – they teared down a M3 with the large battery pack AND premium options (because that are the only ones out there). So a standard one for 35,000$ with the small battery and without the premium options should be even cheaper to make!

+1 That was exactly what I was thinking when I read the story. The base model teardown will likely be thousands lower.

The real challenge at this point is simply putting out the volume to cover the fixed overhead costs, and to get lower price per unit costs on parts from suppliers.

Although they tore down a LR Premium the cost numbers are always adjusted to represent the base model. They want their numbers to be consistent from one OEM to another regardless of which options are on the individual car they tear down.

Should be, but they aren’t building any-they stiffed those customers.

Ok let’s do the math. 10,000/7 days a week = 1428 per day. If they ran the plant 24/7/365 days straight (not possible) that would be 60 cars per hour, or 1 each minute coming off the line. FACT- The same Fremont plant production rate was 8100 per week at peak production when it was run by the super proficient Toyota production management system.

So Tesla just needs to be 25% more efficient to hit 10K/wk? That seems within the realm of possibility.

Any comparison to the old Toyota factory to what Tesla will be doing in the future is simply meaningless now. Tesla has been on a construction bender and is in the process of doubling the Fremont factory.

“Tesla first introduced the expansion plan last year and it included the addition of several new buildings to almost double the size of the already giant electric vehicle factory – 5.3 million square feet to 9.9 million square feet.”



Old style manual build with much less automation … 10k/week is certainly doable. Likely two lines at 5k/week each.

And you know this to be true because of your years in automotive manufacturing, or just your gut feeling and proximity to Silicon Valley?

The Robots were pulled out in April according to Elon. They are saying they are going to host a hackathon to see if they can get robots to work.

they were betting on large set of robots that they could not get to work in q1.
They resorted to an ad hoc restart using human labor, how 20th century. 6 months ago I was thinking they designed the m3 to be built by robots, ir turns out they did not.
Turns out TOYOTA knows what they doing after all, Toyota plants are a mix of human and robots. The battery robots appear to have been much more successful


“Cheap to produce” Isn’t it what Elon said from the beginning?

Yes, and for this Tesla fan it’s very nice to see a teardown analysis that, sharply contrary to the very biased teardown analysis from Detroit specialist Munro, doesn’t try to convince us that Tesla is wasting money on the build, and doesn’t try to convince us that Tesla is clueless about building cars.

I’m sure that Tesla still has a lot to learn in the way of corporate experience in building cars, but the Munro analysis was so hopelessly biased (not to mention showing Munro to be rather clueless about electric cars) that it’s almost entirely worthless.

I don’t know if the analysis highlighted in this article is better informed — I hope so, but don’t have the expertise to judge — but at least it’s not hopelessly biased!

I don’t remember seeing anywhere a break down of Munro’s analysis. Was that published? I thought you had to pay a lot of money to see that? I might have missed it as I don’t check these sites several times a day. Anyhow, since you seem to know, how much did Munro say the Model 3 costs to build?

I’ve never seen a report on costs from Munro’s Model 3 teardown, but a year ago UBS estimated Model 3 costs using Munro’s Chevy Bolt teardown. Those numbers were generally similar, though the German article lacks the detail needed for a good comparison.

There’s also a lot lost in translation. $10k labor is a joke, I guarantee that’s not what the teardown people said.

Too bad Musk botched the Model 3 assembly line design (too much automation) and had to re-tool it to be more like other auto manufacturers.

$28K at 10K units a week is not “cheap” to produce. This does not include operational cost of Tesla which runs about 30% of revenue last Qtr, and also some return on Capital investment. That is 60% of average revenue for “cost of goods sold” ouch…

Making cars isn’t for the faint of heart. Mass producing automobiles requires massive capital investments, and generally yields thin profit margins.

Tesla is the first new American auto maker to make a go of the business in more than 50 years.

Go Tesla!

And their accumulation of massive debt proves how poorly Musk has executed, as he said each car would pay for the next one…….

For a car that is expected to have an average selling price of at least $42,000, $28,000 production cost indeed seems very cheap.

An idea for Tesla to let the rear passengers manage their seats (or equivalents) themselves: An app that let’s them do it as guests to the car

There’s an “app” that humans invented tens of thousands of years ago that lets a back seat passenger request the front seat passenger to do that for them: The app is called “language”. 😉

An “app” which removes human social interaction isn’t a good thing, it’s a bad thing. We’ve already got far too much of that.

While this might be true in general, some people are shy, and prefer to avoid unnecessary communication…

What’s the COGS at today’s 1500/week? Not $28k, that’s for sure.

It’s going to be a very long time until Tesla can get to 10000/week.

The True Believers point to this report as ‘evidence’ that EVs are profitable, but they aren’t. The FACT is that this report points to an estimated cost for a production volume that doesn’t exist yet, and never has.

If EVs were profitable to build, everyone would be doing it; it’s that simple. Tesla is the only one with a chance to be profitable building EVs, and to give the segment have any hope of survival. If Tesla fails, the entire EV market will collapse.

“If Tesla fails, the entire EV market will collapse.”

People are very enthousiastic about e-Golfs, Leafs, Bolts, Ionics, I-Paces. These cars won’t go away if Tesla fails.

There are government incentives to promote EV’s. They won’t go away if Tesla fails.

There are bans on ICE cars looming everywhere. Those won’t be canned if Tesla fails.

If Tesla fails, that will be a major setback, but no collapse.

Ah, the ‘True Believers’. What are you then? A ‘True Hater’? 😉 stop the framing and deal with arguments.

This is just another datapoint that indeed supports the notion that EV’s can be profitable. And a quite strong one. But the proof is in the pudding. Later this year Tesla will definitely put to rest this argument. One way or another.

“In fact, it was revealed that the materials and logistics involved in building the Model 3 add up to about $18,000.”
Very expensive – but propably becasue of the battery that may become cheaper with time.

“Labor costs were determined to be about $10,000. ”
Insanely much!!!!!!

“Labor costs were determined to be about $10,000”

The source article in Wirtschaftswoche says:

“etwa 10.000 Dollar an Produktionskosten pro Auto”

Which is correct? production cost or labour cost?

Let’s also assume 10% overhead (2800 $/car), 10% net profit (2800 $/car) and the 2.5 billion loss (5000 $/car at 500.000 cars/year) per year. If these are added you arrive at 38.600 $/car.

So all you have to do is produce and actually sell 500K/year….That’s a lot. I wonder if that production cost is low enough to offer the car for the sort of (lease)prices that can support a continuous 500K/year sales run.

I am surprised that no one mentioned the Panasonic story. If they make Cobalt free batteries then Tesla’s advantage will be bigger than what it currently is. I could imagine why people are upset and are refuting the idea that the selling price will be $28k and not the cost. Many will feel cheated that they bought a car for $40k or more. The same thing happened with mobile phones. $100 will get you a phone that cost much much more than when mobile first came out. I can not wait for the day that these cars sell for sub $28k with level 5 autonomous driving capability. It will change the world to the better.

How much was added to the production cost to account for depreciation expense? We can’t know the exact number without seeing the books but Tesla has some big loans to pay off and those payments are coming out of sales revenue. It’s a very real factor in determining break-even.

Teslabots don’t figure debt service into profitability, It’s enough to get them to even think Tesla needs to be profitable to sustain itself. Many of them believe an endless supply of $$$ should be available to Musk to “save the world”.