Tesla’s 4680 battery supply chain collapses

[Rippin1Gear]

I mean with the cancelation of the range extender Tesla has to be working on something for us… possible entire battery pack upgrade to the new chemistry. No idea what ever happened to the test mule CT with the “newer” batteries last year.

Sponsored

 

[mitch9]

I mean with the cancelation of the range extender Tesla has to be working on something for us… possible entire battery pack upgrade to the new chemistry. No idea what ever happened to the test mule CT with the “newer” batteries last year.

Yes, probably: "we are shutting down the Cybertruck production line, to make room for Cybercab production and refocus our energy". They could discontinue the Cybertruck, and admit it was a dismal failure, and they are moving on. Other companies have done similar, why not Tesla.

 

[Speedr]

Tesla has been working on dry cathodes for a long time, and recently got it working. I think that explains the reduction of the South Korean nickel cathode material better than they're going to stop ALL 4680 production. Also, there have been plenty of articles that they would start making them in Germany next year...

 

[SCTesla]

Tesla has been working on dry cathodes for a long time, and recently got it working. I think that explains the reduction of the South Korean nickel cathode material better than they're going to stop ALL 4680 production. Also, there have been plenty of articles that they would start making them in Germany next year...

The reduction in the nickel is due to low demand and that's it. The CT is selling 20-25k per year and Tesla thought for sure it would be 125-250k per year.

Plus, there were plans to extend the 4680 to other platforms, but it's really not a better battery than the 2170, still. Everything I have seen show the Semi is still using the 2170s as well.

 

[James Cole]

Wonder when will be able to buy a dry cathode cybertruck? Could the dry battery be much better than current wet 4680?

 

[henchman24]

Wonder when will be able to buy a dry cathode cybertruck? Could the dry battery be much better than current wet 4680?

Dry vs wet doesn't have a real impact at the consumer level. It is more about cost. Not something that people should really be looking to 'upgrade' to. It is something shareholders should care a out.

The next jump for the consumer level in 4680s is the addition of silicon carbide into batteries. Depending how much ends up in the cells, it could represent a very small improvement (~5% range, slightly better charging) to a large ~20% improvement. Rumors are more around the 10-15% with better charging curve (think 250 kW at 40% instead of 25-28%). This likely makes it into the CT first in mid-to-late 26 for a refresh and some public testing (Semi is likely too). The assumption is that goes well, then the 4680 Y will quickly return as there will be enough energy density to pack in an ~85kWh pack at that point, and this is where the Germany investment numbers are stemming from.

The Cybercab batteries are supposed to be 4680s, but different than CT/Semi/Y versions planned. LFP chemistry and much, much lower energy density/power requirements.

The CT has also already had a battery change happen. It happened late summer/early fall with the move to 973 chemistry vs 955. Practically this has little consumer impact. Slightly wider temperature range, so slightly better charging and cold weather performance.

Now the whole article is Fred being Fred.

 

[James Cole]

very informative thank you…!

About the actual battery change in the CT you mentioned in summer/early fall is it this year and happen to know around what vin approximately?

JC

Dry vs wet doesn't have a real impact at the consumer level. It is more about cost. Not something that people should really be looking to 'upgrade' to. It is something shareholders should care a out.

The next jump for the consumer level in 4680s is the addition of silicon carbide into batteries. Depending how much ends up in the cells, it could represent a very small improvement (~5% range, slightly better charging) to a large ~20% improvement. Rumors are more around the 10-15% with better charging curve (think 250 kW at 40% instead of 25-28%). This likely makes it into the CT first in mid-to-late 26 for a refresh and some public testing (Semi is likely too). The assumption is that goes well, then the 4680 Y will quickly return as there will be enough energy density to pack in an ~85kWh pack at that point, and this is where the Germany investment numbers are stemming from.

The Cybercab batteries are supposed to be 4680s, but different than CT/Semi/Y versions planned. LFP chemistry and much, much lower energy density/power requirements.

The CT has also already had a battery change happen. It happened late summer/early fall with the move to 973 chemistry vs 955. Practically this has little consumer impact. Slightly wider temperature range, so slightly better charging and cold weather performance.

Now the whole article is Fred being Fred.

 

[Nine_block]

Dry vs wet doesn't have a real impact at the consumer level. It is more about cost. Not something that people should really be looking to 'upgrade' to. It is something shareholders should care a out.

The next jump for the consumer level in 4680s is the addition of silicon carbide into batteries. Depending how much ends up in the cells, it could represent a very small improvement (~5% range, slightly better charging) to a large ~20% improvement. Rumors are more around the 10-15% with better charging curve (think 250 kW at 40% instead of 25-28%). This likely makes it into the CT first in mid-to-late 26 for a refresh and some public testing (Semi is likely too). The assumption is that goes well, then the 4680 Y will quickly return as there will be enough energy density to pack in an ~85kWh pack at that point, and this is where the Germany investment numbers are stemming from.

The Cybercab batteries are supposed to be 4680s, but different than CT/Semi/Y versions planned. LFP chemistry and much, much lower energy density/power requirements.

The CT has also already had a battery change happen. It happened late summer/early fall with the move to 973 chemistry vs 955. Practically this has little consumer impact. Slightly wider temperature range, so slightly better charging and cold weather performance.

Now the whole article is Fred being Fred.

This is perfectly summarized. I like battery tech and keep up with this space and this is dead on the money. The author of the linked article is a drama queen and the worst part of that website. I rather like the other authors there.

 

[freddms]

I know it’s Electrek but I found the article interesting.

https://electrek.co/2025/12/29/tesla-4680-battery-supply-chain-collapses-partner-writes-down-dea/

It's just annoying to read these articles. The death of the Cybercab implied before it's even released. You can't make this garbage up.

The truck is not selling well, but other manufactures have completely cancelled their EV's. Meanwhile EV sales continue to take a bigger percent of total car sales. Let's give it some time. My CT is incredibly good - except for a camera visibility problem (corrected under warranty) I have had zero problems. Did have my GPS glitch for an afternoon. Other than that I'm 98% FSD while driving, and it's a hell of a truck that puts all other to shame (except for pulling a boat/trailer). After driving three full size vans, and 3 pickup trucks through my life, I will buy another if they are available.

From the article:
"The ‘Cybercab’ is also supposed to be using the 4680 cells, but we will have to wait and see how that goes.

It’s also a vehicle program that could go the way of the Cybertruck. CEO Elon Musk is insisting that it will launch in early 2026 without a steering wheel, but Tesla has yet to solve level 4 autonomous driving."

 

[henchman24]

very informative thank you…!

About the actual battery change in the CT you mentioned in summer/early fall is it this year and happen to know around what vin approximately?

JC

Until someone tears down a few, we are unlikely to fully know. Somewhere between July and September 25 a change over happened. Practically the difference is minimal (much, much less than the difference between 3/Y LG and Panasonic cell vehicles). The capacity difference is next to nothing (maybe 2%). The temperature window difference does allow some extra charging temperature headroom (though it doesn't seem the curve changed much, so far), but just lessens the occurrence of the snowflake.

 

[Scorpious]

People seem to forget they are using the 4680 for the semi and about to ramp that up. This is just a secondary supplier they don't need at the moment. Another case of semi true fake news to blow up a standard business practice called managing suppliers.

 

[SCTesla]

People seem to forget they are using the 4680 for the semi and about to ramp that up. This is just a secondary supplier they don't need at the moment. Another case of semi true fake news to blow up a standard business practice called managing suppliers.

Semi is still using 2170s.

 

[henchman24]

People seem to forget they are using the 4680 for the semi and about to ramp that up. This is just a secondary supplier they don't need at the moment. Another case of semi true fake news to blow up a standard business practice called managing suppliers.

Slight correction, currently the Semi is using 2170s. The 'next gen' version that is going to ramp up, is designed around 4680s. Odds are we see a bit of a mix for a little bit before they transition over fully to 4680s, but the Semi program will be 4680 long-term.

The 4680 format is becoming standardized around the industry. There are now major battery manufacturers making 4680 cells. It isn't just Tesla. It has even been expanded to 4695 and 46120 formats as well (there are others too, these are just the popular lengths). Frankly, I think people forget that the cell is simply just a format. To counteract some of the issues of the cell design you need certain technologies (namely tabless), but all 18650, and 2170 technology can simply swap over now if you're willing to accept lesser density. Tesla in their program is trying to add in a few extra technologies and this is the fundamental production issue they have had. They can simply go buy a standard wet, tabless 4680 (or 4695/46120) cell from CATL, LG, BYD or Samsung today.

 

[SCTesla]

Slight correction, currently the Semi is using 2170s. The 'next gen' version that is going to ramp up, is designed around 4680s. Odds are we see a bit of a mix for a little bit before they transition over fully to 4680s, but the Semi program will be 4680 long-term.

The 4680 format is becoming standardized around the industry. There are now major battery manufacturers making 4680 cells. It isn't just Tesla. It has even been expanded to 4695 and 46120 formats as well (there are others too, these are just the popular lengths). Frankly, I think people forget that the cell is simply just a format. To counteract some of the issues of the cell design you need certain technologies (namely tabless), but all 18650, and 2170 technology can simply swap over now if you're willing to accept lesser density. Tesla in their program is trying to add in a few extra technologies and this is the fundamental production issue they have had. They can simply go buy a standard wet, tabless 4680 (or 4695/46120) cell from CATL, LG, BYD or Samsung today.

Yes.

Elon has already said they will be using outside 4680s. Battery manufacturing is difficult and they haven't seen the cost savings or gains they hoped.

 

[Jhodgesatmb]

The company that is a supplier wrote down its investment by 99%. If Tesla doesn't need cathodes, then it doesn't need batteries. I would say that matches the 99% over hyped of all the millions of CT orders. Was just a matter of accounting. Not really news. We all know the CT doesn't sell well.

All it means, if true, is that Tesla is getting cathode materials elsewhere. To stretch this information to 4680 production is an unreasonable prediction.

Sponsored