Tesla's Earnings Call Recap - HW3 Upgrade, Unsupervised FSD, FSD V14 and More

Not a Tesla App

If you missed Tesla’s 2024 Q4 earnings call or just want to review everything that was shared, don’t worry - we’ve got you covered. Below is a list of all the key points made during the call, which includes the popular Q&A session with Elon Musk and other executives. We broke down the information into several categories for easier reading.

If you prefer, you can also listen to Tesla’s earnings call on demand.

Overall Finances

• Tesla’s total revenue increased 2% year-over-year in Q4, to $25.7B.

  • Tesla saw growth in Energy Generation, Storage, Services, and Other product/service groups

  • Tesla saw growth in vehicle deliveries

  • Tesla had higher regulatory credit revenue (likely due to pressures from credits ending soon)

  • Reduced Model S, 3, X, and Y average sales price (ASP), leading to increased sales

    • Attractive financing options, including 0% financing, spurred additional sales

• Tesla’s operating income decreased 23% YoY in Q4, down to $1.6B

  • That is an overall margin of 6.2%

  • Reduced vehicle ASP impacted margins

  • Increase in operating expenses driven by AI and other R&D projects

  • Income increased in:

    • Energy Generation, Storage, Services, and Other categories

    • Lower cost per vehicle, including lower raw material costs

    • Higher regulatory credit revenue

• Tesla has $36.6B in free cash, cash equivalents, and investments. This is an increase of $2.9B over last year, primarily the result of $2B in positive cash flows.

Production, Revenue, COGS

• Energy storage production has grown YoY since 2023

  • Growth came from both Megapack and Powerwall

  • Growth is supply-constrained - Tesla can’t produce them fast enough - both residential and commercial customers are ordering quickly

• Tesla delivered over 2 million vehicles in 2024, a new record for production and deliveries (likely includes used or other types of vehicles since total new vehicle deliveries were at almost 1.8m in 2024)

  • The lowest amount of un-delivered vehicles this year ever, very few vehicles sitting on service center/delivery lots

  • The Model Y was once again the best-selling vehicle of 2024 - of any kind

  • Tesla’s current constraint for production is battery packs

    • Making progress in addressing that constraint

    • Progress should improve drastically next year and the years after for battery pack production

• Tesla grew in volume in China greatly, even though it faces extreme competition in the country

• Overall cost per car is down to below $35,000 per vehicle.

• All factories will start producing the new Model Y next month

  • This rapid and unprecedented change could result in an impact on production

  • Margins will be impacted due to launch-related costs for Q1 and Q2

• The referral program had an impact on cost and will continue to do so as the program continues

• Tesla doesn’t expect margins for service, but service remains GAAP-positive

New Products

• Several new products in 2025

  • 1st half of 2025 - More affordable model is still coming

  • Per dollar basis, Tesla has the most compelling lineup

    • Best software in the business, capable of autonomy

    • Prices comparable to the mass market, products compared to premium competitors

• Optimus could potentially launch for commercial use in 2026.

Looking Forward

• Tariffs

  • Uncertainty, focusing on localizing supply chains

  • Tariffs are very likely, will have an impact on business and profitability

• AI is the biggest driver in cost

• Tesla increased CapEx by $2.4B

  • Focused on improving CapEx efficiency, using a targeting manner to get immediate benefits

  • Build out of Cortex accelerated the rollout of FSD V13

  • $5B in AI-related CapEx

  • 2025 CapEx expected to be flat

  • Focus of critical investments will be on manufacturing, AI, and robotics.

    • These are expected to bear fruit in 2026, 2027, and 2028.

• 2025 will be a pivotal year for Tesla

  • Lots of investments continue to be made

  • Setting the pace for the next phase of growth

  • Tesla is focusing on maximizing volumes and doubling down (or 10x) on Autonomy and real-world AI.

• Elon sees a path for Tesla to be the most valuable company in the world by far - it is a difficult but achievable path

  • The path means Tesla will be worth more than the next top 5 companies combined

  • This will be due to autonomous vehicles and autonomous robots

    • Tesla began laying the groundwork in 2024

    • Tesla will begin building the manufacturing lines in 2025

    • “Epic 2026 and ridiculously good 2027 and 2028” - Musk

  • The future will be coming very fast, the inflection point for human history (autonomy/AI)

FSD and Autonomy

• Tesla has achieved exponential progress in FSD

  • People don’t realize how good FSD is, and many people have no experience with it

  • Those with previous experience need to try it again - it has improved drastically over the last year

    • Last year’s experience was with a toddler - this year’s experience is with a grown-up

• Passenger cars have 10 hours of utility per week - out of 168 hours a week.

  • An autonomous car would be of utility for 55 hours a week - about 1/3rd of the week

  • Autonomous vehicles are useful for both cargo and people delivery

  • No incremental cost change, just a software update to enable 5x utility

  • Largest asset-value increase in human history

  • Bigger than anything else in its history - bigger than the first car, the roadster, the Model 3, the Model Y

  • Launch of Unsupervised FSD will be a turning point - true, real-world AI that actually works

  • No company in the world is as good as Tesla at real-world AI

FSD Going Global

• FSD works very well in the US, but over time it will work just as well everywhere else

  • Europe is a challenge due to regulation

    • Netherlands presents FSD to EU in May, expected to be approved EU-wide then. Won’t happen sooner than May

    • Unsupervised FSD in Europe could be even further away

  • China is another challenge - won’t allow training videos to leave China, and US won’t let training be done in China

    • Tesla is using public videos of Chinese streets on the internet to train Chinese FSD

    • Bus lanes in China are one of the biggest challenges there due to restrictions based on time, etc - automatic tickets for being in the wrong lane in China

• Q4 Safety Report

  • Continual year-over-year safety improvements for those using FSD

  • People go to manual driving to check their phone so that they don’t strike out/get beeped at - and then go back to FSD

    • Significantly less safe than just being on FSD

  • Tesla is getting to the point where FSD is an order of magnitude safer - already 8.5x safer than a human driver

FSD V14

• FSD V14 will be another significant step

  • Cortex launched at Giga Texas, which is a significant help in training FSD, and Tesla will continue expanding it

  • Tesla will need 10x the current compute to make Optimus fully useable (a thousand times more complex and a thousand times more use cases than FSD)

  • Tesla will spend the money on training compute over time, it won’t suddenly spend $500B on compute tomorrow

  • The cost of training is dropping dramatically with time

• FSD V14 will focus on improving neural nets, working with auto-regressive transformers

  • Model size and context size will continue to grow and scale up

  • Context size is restricted by on-board memory, and will require optimization

  • Audio support to be further expanded and become relevant in V14

  • Data from tricky edge cases gathered by the fleet will help V14 make better decisions

Unsupervised FSD

• Unsupervised FSD in Austin, Texas, in June 2025, as a paid robotaxi service.

  • Already happening at Fremont, will be happening at Austin soon, and other factories later

    • Cars know what lane they need to go to for pickup/delivery

    • They drive from the line to their spot and do it reliably every day, thousands of times a day

    • Teslas will be in the wild, with no driver, as of June

  • Tesla’s solution is generalized - doesn’t require high-precision maps.

    • Tesla is starting with Austin to make sure everything is ok and to test the waters, to ensure the safety of the general public and those in the cars

  • More cities will come in the future

    • Everywhere in North America by next year

  • Unsupervised FSD is limited by regulatory issues, not technical capability.

  • Tesla’s fleet will be using Unsupervised - you won’t be able to bring your own vehicle to Austin and enable it just yet.

    • Robotaxi fleet will be available to be ride hailed and Tesla will charge for it.

    • Tesla is still working on the experience - the app, arrival, payment, billing, exiting the vehicle, etc.

    • In 2026, owners will be able to add their vehicles to the Unsupervised FSD fleet.

    • Tesla needs to be confident that the probability of injury or accident is extremely low before they allow eyes-off / Unsupervised FSD on customer vehicles

• Unsupervised FSD will also happen in California this year - in fact, in many regions of the US this year.

  • Austin is Tesla dipping their toes into the water to confirm it works and is safe.

  • Tesla is aiming for a safety level significantly above the average human driver

    • Standard is very high, if there is even one accident, it will get worldwide headlines, even though 40,000 people die in car accidents every year

  • Hold-back is an excess of caution

• How many times did you have to intervene for definite safety reasons?

  • Don’t compare to feeling like you had to intervene - Tesla is close to eliminating critical safety interventions being normal. They’re already very rare.

Hardware 3 - Replacement Confirmed

• Elon confirmed HW3 will need to be replaced - it will be replaced for free for those who previously purchased FSD outright.

  • This will be expensive - but it won’t have a massive impact because less people purchased it outright versus subscribing

• Tesla still working on HW3 in the meantime - V12.6 is a baby V13 and a significant improvement over V12.5.4.2

• HW3 releases will trail AI4 releases for the foreseeable future

Licensing FSD

• Tesla is in discussions with multiple major car manufacturers for licensing FSD

• Tesla is advising companies to take apart their vehicles, and look at requirements

  • Cameras, compute, thermal limitations, capabilities and design

• Tesla will only sign if the volume is very high - otherwise, it isn’t worth it for Tesla due to engineering constraints

  • Cybertruck’s FSD limitations are a direct result of this

• Interest in licensing FSD will become even higher once Unsupervised FSD becomes proven

Optimus

• It is difficult to predict when Optimus will arrive. It is not design-locked for 2025

  • Tesla is building the tracks, re-designing the train, and designing the station in real-time while putting it all together

  • Tesla used all custom parts for Optimus because they couldn’t find anything that met their requirements

  • Constantly improving Optimus - evolving in a good direction

• The internal plan is to build 10,000 by the end of 2025 - but that’s optimistic

  • Tesla will build several thousand at least

  • They will do useful things by the end of the year - at least within Tesla’s factories.

• This is “production design 1”, and use in Tesla’s factories will inform “production design 2” for next year.

• Optimus will ramp faster than anything else they’ve ramped - they’ve previously done half an order of magnitude a year - they want to aim for one full order of magnitude per year

  • 100m per year in the far future

• Most sophisticated humanoid hand

  • Optimus will be able to play piano, thread a needle

  • Extreme level of precision with the hand that no other competitor has gotten close to

  • Other companies can’t produce thousands or millions of units - they can build one-off hand-build units or a few prototypes

  • Other companies have extremely limited real-world experience and data

• Long term, Optimus can be $10-20B in revenue

• More compute means more Optimus, more Optimus means more compute

• Tesla doesn’t intend to begin Optimus sales just yet - Elon was careful not to answer questions related to sales and pricing

  • Tesla wants to use Optimus immediately for the boring/tedious/dangerous tasks around its own factory first

  • This will help training as well, close the loop for improvement

  • Production Version 2 - launches middle of next year

    • Production line to build up to 10K units a month, rather than the current limited production

    • This could be the version that is potentially sold to companies in the second half of next year

    • Likely won’t be sold directly to retail customers

  • The expected ramp is really fast

  • Demand will not be a problem, even at a high price

  • At 1m units of production per year, the production cost will be under $20,000 per unit

    • This is likely several years away

  • Price will be set by market demand

Semi

• Semi Factory just had its roof finished last week

• Mechanical equipment will be installed in the next couple of months

• Late this year comes the high-volume variant of Semi

• Scale production begins end-2025/early-2026

• Will meaningfully contribute to Tesla’s revenue next year

• There is a shortage of truck drivers in America, and people are humans - they get tired and need breaks

  • It is a tough job, and lots of people are leaving the profession rather than joining it.

  • Future logistics problem - autonomy will help to bridge the gap and meet the need for trucking

• Semi will likely be a $12B/year endeavor in the future once the ramp is complete

Energy

• Tesla sees this as a “down-to-Earth” topic - it's real, and it’s happening now

• Storage is a big deal and will become even more important in the future

  • Total demand for electricity storage will grow massively over time

  • Enables far greater energy output to the grid than is currently possible

  • Vast majority of the global grid doesn’t have storage - power plants are designed for high peaks rather than steady energy

  • Lots of storage will help to increase the energy available - by up to 2x what is available now

• Lots of demand for stationary storage, both residential and commercial

• Mega Shanghai is going fast

  • Elon confirms a 3rd Megafactory for stationary storage, but not the location

  • Tesla will NOT rob from vehicle storage to feed into stationary storage, nor vice versa

• Tesla still working on Solar Roof, haven’t given up

  • Core part of the residential product portfolio

  • Is a premium product - like the Model S or Model X - and priced like it

  • New versions of the Solar Roof are being designed and worked on still

  • Best time to put in a solar roof is either on a new home or a house needs a new roof - solar roof makes sense then

  • Combined with Powerwall, it makes owners self-sufficient, even if the grid is off for several days

Other Topics

• LiDAR is still a crutch

  • Tesla has radars in the Model S and Model X - they ship turned off

  • Humans drive without shooting lasers from their eyes - they drive with eyes and a brain.

    • Digital equivalent is cameras and a neural net

    • Road system is designed for this, using LiDAR doesn’t help the generalized solution

    • LiDAR is good for a specific solution - which has to be trained neighborhood by neighborhood

    • LiDAR doesn’t work in fog, doesn’t work in inclement weather, and doesn’t make sense on roads

    • SpaceX uses LiDAR, but it's the wrong solution for cars on roads that were made for people

• Real-world AI and the US, next 4 years

  • Talented Americans need to look at making manufacturing cool again in America

  • Move from law, finance, etc, to manufacturing. Too much talent focused on there

• Tesla’s mission of electrifying the world through sustainable transport and energy

  • It’s inevitable that all transport will eventually be autonomous and electric, including aircraft

    • Loss of rebates or changes in policy won’t change that

  • Can’t be stopped, it will eventually happen, just like the internal combustion engine taking over

That’s the complete step-by-step replay of the entire earnings call. It’s a lengthy one, but there was a lot of information shared across all of Tesla’s pillars. There’s a lot to be excited about, from FSD to the next-gen vehicle, Optimus, the Robotaxi network, and more. These next few years will be huge for Tesla and we hope you’ll join us for the ride.

Follow

Follow

Follow

Tesla’s Hollywood Diner: In-Car Controls & Theater Screens Turned On [VIDEO]

Sky Fox

Tesla has been quietly building something pretty unique on the streets of Hollywood—and we’ve been keeping a close eye on it. That’s right: the long-anticipated Tesla Hollywood Drive-In, Diner, and Supercharger is shaping up to be one of the most creative approaches to EV charging yet.

But this isn’t your average Supercharger site. It features a retro-futuristic, 24-hour diner and an outdoor movie theater, blending the charm of the 1980s with the tech-forward vibe of the 2020s.

Construction has been underway for quite some time, with the Cybertruck-inspired stainless steel structure going up around mid-2024. And more recently, we spotted hints in the Tesla app that the Diner’s debut is getting close.

Diner Controls in App / Car

Tesla hacker Greentheonly managed to spot some new evidence of changes within Tesla’s 2025.8.6 software build, which includes a new modal for “Charger Diner Controls”. While there isn’t anything to activate this new modal just yet, it is an indication of how Tesla intends to integrate the experience right into each and every Tesla.

The text in the modal reads: “Plug in, place your order, recharge” - which means that once you arrive and are plugged in, you’ll be able to place your order - and likely pay through Tesla’s payment system. It seems like this will all be done through either the Tesla app or the vehicle’s tocuh screen. No need to fumble with a separate app or pull out your wallet.

This is definitely a level of integration you just can’t have anywhere else - and just like any other Supercharger site, the Tesla Diner will appear as a suggested site for charging when you’re navigating or looking around the Hollywood area.

This is our best look yet at the two 45-foot LED screens at @Tesla's upcoming Hollywood diner and Supercharger.

We can also see the second Supercharger lot on the left with solar canopies. This will be one of the largest Supercharging spots in North America. pic.twitter.com/aWUe67M5vw

— Sawyer Merritt (@SawyerMerritt) April 18, 2025

45-Foot Screens Ready

Those two absolutely gigantic 45-foot LED screens have now been turned on for the first time, meaning that Tesla is likely close to opening the Supercharger site. The screens were spotted displaying some test patterns, as well as some Tesla-themed content. It really brings the construction site to life - and it's the clearest sign that the Diner is well on its way to completion.

We’re hoping that this first Tesla Diner sees plenty of success - and that Tesla begins opening more of these types of facilities throughout North America, and eventually globally. Given Tesla’s penchant for unique Superchargers, we imagine this could truly be the case one day.

Follow

Follow

Follow

Tesla’s FSD V13 Pushes HW4 Hardware Capabilities; End of Line for HW3?

Not a Tesla App

As Tesla continues to push the capabilities of its hardware as it prepares for autonomy with FSD Unsupervised, there has been a distinct divergence in features between HW4-equipped vehicles on FSD V13, and HW3-equipped vehicles on FSD V12.6. 

There are definitely reasons for this, and Tesla hacker Greentheonly, over on X, went into great detail on some of the reasons why this divergence has continued and is likely to continue.

If you’re interested in some background, we recommend the following articles. Although they’re not necessary, you may enjoy reading them before diving right in if you enjoy this topic.

• How FSD Works - Part 1

• How FSD Works - Part 2

• How Tesla Optimizes FSD

• Tesla’s Universal AI Translator

Back in the summer of 2024, we reported that Tesla’s HW4 had yet to reach its full potential, and that HW3 still had some compute to work with. However, FSD has undergone a rapid evolution - there are now fundamental differences driven by the hardware capabilities and sheer scale of the latest builds of FSD.

FSD V13’s Gargantuan Neural Networks

FSD V13 is a substantial leap for Tesla. It features substantially larger neural networks - the core brain that processes all the environmental data and makes decisions. Those networks are far larger in V13 than in V12, which results in a much smoother and human-like expereience.

Tesla’s FSD computers contain two nodes. According to Green’s analysis of FSD V13, Node B, which handles the actual end-to-end driving logic, tripled in size from 2.3 GB in V12 to a staggering 7.5 GB in V13. That’s a substantial increase, and one of the main reasons why HW3 can’t handle FSD V13. While Interestingly, Node A on HW4 remains similar in size to FSD V12, using up about 2.3 GB.

However, this massive size is also pushing the limits of HW4. While this isn’t necessarily a big problem immediately - at the end of the day, there is a limited amount of compute and ram available and model sizes will likely grow in the future. In fact, Tesla is planning to scale model sizes in the next FSD version and has already said they’ll need to optimize them, as context size is limited by on-board memory.

HW3 Limitations

Since HW3 can’t run the latest FSD models, where does this leave millions of Tesla vehicles equipped with the now-legacy FSD hardware? Are they hitting a dead end? Not exactly - Tesla has promised to replace HW3 computers with new units during the Q4 2024 Earnings Call, but only for those who have purchased FSD and only if Tesla is no longer able to update HW3 vehicles.

There’s no doubt Tesla hopes to achieve autonomy and then work backwards to make it compatible with HW3, but that’s unlikely to happen with the limited ram available on HW3.

Tesla’s AI team has continued to say that they’re still working on updating HW3 vehicles - but that they will be taking a backburner to updates for HW4 until the team can optimize the model efficiently enough to run on the smaller footprint. That means V12.6 isn’t just V13 “lite” - it is a distinctly different software branch tailored specifically for HW3’s capabilities.

Green's analysis showed that HW3 under V12.6 runs NNs totaling 1.2 GB on Node A and 3.1 GB on Node B, which is significantly smaller than the footprint of V13. Back in November, we discussed how Tesla is running a smaller and optimized FSD model on HW3, with the specific goal of achieving similar performance to V13. FSD V12.6 appears to be the current implementation of that smaller model strategy.

Green also found that of the total neural nets used for FSD, 135 are shared between the current V12.6 and FSD V13.2 releases, which is the result of Tesla’s AI team working to optimize and refine the V13 model for HW3. So some neural nets are shared between the two sets of hardware, while others are specific to HW3 or HW4.

AI4 in the Future

The fact that V13 is already stretching HW4’s limits makes us wonder about Tesla’s imminent release of AI5. Last year, Tesla mentioned that AI5 would be 18 months away - which is now about 8 months away, and potentially even sooner to help solve the autonomy problem.

At the same time, Tesla is looking to launch Unsupervised FSD in June in Austin, Texas, with presumably HW4, which would indicate that they’re confident in its capabilities.

Unless Tesla launches the Robotaxi network with AI5, we don’t believe AI5 will be a “must have” item. Undoubtedly, it will result in similar improvements to HW4 compared to HW3. However, the software needed to take advantage of the newer hardware could still be years away.

All eyes will be on Tesla’s Robotaxi network launch and the hardware required for it. For now, HW4 owners are in good hands with future updates are on the horizon, and we’re certainly excited for what’s next.

Follow

Follow

Follow