Tesla

Tesla, Inc.
Type	Public
Traded as	Nasdaq: TSLA S&P 500 component
Industry	Automotive Energy storage Robotics Artificial intelligence
Founded	1 July 2003
Founder	Martin Eberhard Marc Tarpenning (with Elon Musk, JB Straubel and Ian Wright as co-founders by agreement)
Headquarters	Austin, Texas, United States
Key people	Elon Musk (Chief executive officer) Robyn Denholm (Chair)
Products	Model 3, Model Y, Cybertruck, Semi, Powerwall, Megapack, Solar Roof, Optimus
Employees	Approximately 125,000
Website	tesla.com

Tesla, Inc. is an American company that makes electric vehicles, battery energy storage systems, solar products and, increasingly, software and robotics. It is headquartered in Austin, Texas, having moved from Palo Alto, California, in 2021.

The company was incorporated on 1 July 2003 by Martin Eberhard and Marc Tarpenning, and Elon Musk, JB Straubel and Ian Wright are recognised as co-founders under a 2009 settlement. Musk led the Series A financing in February 2004, became chairman, and has been chief executive since 2008.

Tesla is the company principally responsible for the fact that the automotive industry is electrifying. Before it, the electric car was a compliance exercise producing small, slow and unattractive vehicles; Tesla built expensive fast ones that people wanted, used the proceeds to build cheaper ones, and by demonstrating demand forced every established manufacturer to respond.

Its market capitalisation exceeded US$1 trillion between November 2024 and February 2025 and again from May 2025 to July 2026, a valuation that has never been justifiable by automotive earnings alone and that rests on expectations about autonomous driving, energy storage and robotics.

In November 2025 shareholders approved a compensation package for Musk valued at approximately US$1 trillion, contingent on operational and market capitalisation milestones, the largest ever awarded.

In January 2026 Tesla lost the position of largest battery electric vehicle manufacturer to BYD of China, which reported 2,256,714 battery electric deliveries for 2025 against Tesla's 1,636,129.

On 28 January 2026 Musk stated that the Model S and Model X would be discontinued in the second quarter of that year to concentrate manufacturing capacity on the Optimus humanoid robot.

History

Founding, 2003 to 2008

Martin Eberhard and Marc Tarpenning had founded and sold NuvoMedia, an electronic book company. Eberhard's interest in electric vehicles arose from a concern about oil dependence and from the observation that the lithium-ion cells used in laptops had improved to the point of being usable in a car.

They incorporated Tesla Motors on 1 July 2003, named after Nikola Tesla, whose alternating current motor design the company's own motors descend from. Ian Wright joined shortly after.

The founding thesis was specific and correct. Electric vehicles had failed commercially because they were built as inferior versions of ordinary cars. Instead, a company should begin with an expensive sports car, where customers pay for performance and where low volume suits a startup; use the proceeds to build a cheaper saloon; and use those proceeds to build a mass-market vehicle. Musk published this in 2006 as the Secret Master Plan.

The company raised US$7.5 million in Series A in February 2004, of which Musk provided US$6.5 million, becoming chairman. JB Straubel joined in May 2004 as chief technology officer and was responsible for the battery architecture, which used thousands of small cylindrical laptop cells rather than large purpose-built ones, a decision that gave Tesla access to a mature and cheap supply chain when nobody made automotive cells at scale.

The Roadster, based on a Lotus Elise chassis, was revealed in 2006 and delivered from 2008. It accelerated to 60 miles per hour in under four seconds, had a range above 200 miles, and cost approximately US$109,000. Approximately 2,450 were sold.

Its development nearly destroyed the company. Costs exceeded estimates by a very large margin, the transmission did not work and was eventually eliminated, and the 2008 financial crisis closed the capital markets. Eberhard was removed as chief executive in 2007 and left, and subsequent litigation between him and Musk over the circumstances and over who had founded the company was settled in 2009 with the five-founder agreement.

Musk became chief executive in October 2008 with the company days from insolvency, reportedly borrowing personally to meet payroll. A US$50 million investment from Daimler in May 2009 and a US$465 million loan from the United States Department of Energy under the Advanced Technology Vehicles Manufacturing programme, approved in January 2010, provided the capital. Tesla listed on Nasdaq on 29 June 2010 at US$17 a share, the first American car company to go public since Ford in 1956, and repaid the government loan in 2013, nine years early.

Model S and the demonstration, 2012 to 2016

The Model S, delivered from June 2012, is the vehicle that changed the industry's assessment.

It was designed as an electric car rather than adapted from one, with the battery as a flat structural floor, which lowered the centre of gravity and produced handling that petrol saloons could not match; two boots, since there is no engine; a 17-inch touchscreen replacing almost all controls; and over-the-air software updates, which no car had.

Motor Trend named it Car of the Year in 2013 by unanimous vote, the first electric vehicle to win, and Consumer Reports scored it 99 out of 100 and subsequently 103, which required the magazine to explain its scale.

The Supercharger network, opened from 2012 and provided free to Model S buyers initially, addressed the objection that an electric car cannot be driven between cities. Building the charging infrastructure was an expense no car manufacturer had contemplated and is among Tesla's more consequential decisions.

The Model X, delivered from 2015, added falcon-wing doors whose complexity delayed production and damaged quality, and which Musk has described as an error.

Autopilot was introduced in October 2014 with hardware and activated by software update in October 2015, providing lane keeping and adaptive cruise control, and was the first such system delivered to a mass market vehicle by software update.

Gigafactory Nevada, announced in 2014 as a joint venture with Panasonic, was intended to produce more lithium-ion cells than the entire world output at the time, on the correct judgement that battery supply would be the constraint on the industry.

Model 3 and production hell, 2016 to 2020

The Model 3, revealed in March 2016, received approximately 400,000 reservations at US$1,000 each within weeks, which was both a demonstration of demand and a very large obligation.

Production, beginning in July 2017, was catastrophic. Musk had attempted to automate the assembly line to a degree nobody had achieved, and the automation did not work, a failure he acknowledged publicly with the observation that excessive automation was a mistake and that humans are underrated. Output ran at a fraction of the target for many months, the company lost approximately US$700 million in a quarter, and Musk slept at the factory.

The target of 5,000 cars a week was reached in June 2018 only by building an additional line in a tent in the car park, erected in approximately three weeks. Tesla was within weeks of insolvency during this period, a fact Musk confirmed subsequently.

The same period produced the funding secured episode. On 7 August 2018 Musk stated on Twitter that he was considering taking Tesla private at US$420 a share with funding secured. The funding was not secured. The Securities and Exchange Commission sued for securities fraud; the settlement in September 2018 required Musk to pay US$20 million personally, the company to pay US$20 million, Musk to cease being chairman for three years, and his statements about the company to be reviewed by a lawyer before publication, an arrangement he has since challenged repeatedly.

From 2019 the company became profitable and remained so. Gigafactory Shanghai, begun in January 2019 and producing within eleven months, was the first foreign-owned car plant in China permitted without a joint venture, and it became Tesla's largest and most efficient facility.

The Model Y, delivered from 2020, became the best-selling vehicle of any kind in the world in 2023.

Peak and reversal, 2020 to 2026

Tesla's shares rose approximately eightfold in 2020 and it was added to the S&P 500 in December of that year, at the time the largest company ever added.

The company opened Gigafactory Berlin and Gigafactory Texas in 2022, moved its headquarters to Austin in 2021, and reached annual deliveries of approximately 1.8 million vehicles in 2023.

Several difficulties then arrived together.

Competition. BYD and the other Chinese manufacturers reached price and quality levels Tesla could not match in China, and began exporting. Every established manufacturer had electric vehicles by 2023, and while most were unprofitable they took share.

Price cuts. Tesla reduced prices repeatedly from late 2022 to sustain volume, which reduced margins from approximately 30 per cent gross on automotive to roughly half that, damaged residual values and angered recent buyers.

The Cybertruck, delivered from November 2023, sold far below the approximately one million reservations taken, suffered repeated recalls, and its stainless steel construction and angular form, which Musk had presented as a deliberate provocation, limited its market.

Product ageing. The Model 3 and Model Y received updates and the range has otherwise been static for years, and the promised US$25,000 vehicle was repeatedly delayed and then reframed as a cheaper version of existing models.

Musk's politics. Musk's acquisition of Twitter in October 2022, his subsequent statements, his endorsement of Donald Trump in 2024 and his role in the Department of Government Efficiency during 2025 produced a measurable commercial effect. European sales fell sharply, with German registrations down by very large proportions and British registrations falling more than 29 per cent year on year in December 2025. Polling in several countries found Tesla's brand favourability among the demographic most likely to buy an electric vehicle had fallen substantially, and protests including vandalism of vehicles and showrooms occurred in several countries. By February 2025 the share price had fallen sharply and the cause was widely attributed to this.

Deliveries fell in 2024 for the first time in the company's history and fell again in 2025.

In January 2026 BYD overtook Tesla as the largest battery electric vehicle manufacturer, reporting 2,256,714 deliveries against Tesla's 1,636,129.

The pivot to artificial intelligence, 2024 to 2026

Musk's response has been to reframe Tesla as an artificial intelligence and robotics company rather than a car manufacturer.

Robotaxi. Tesla unveiled the Cybercab, a two-seat vehicle without steering wheel or pedals, in October 2024, and began a limited supervised robotaxi service in Austin in June 2025 with safety monitors in the vehicles, extending to other cities subsequently. Musk has forecast full autonomy annually since 2016.

Optimus. The humanoid robot, announced in 2021 with a person in a costume and demonstrated in progressively more capable prototypes since, is the product Musk has described as eventually the most valuable the company makes. On 28 January 2026 he stated that the Model S and Model X would be discontinued in the second quarter to free manufacturing capacity for it.

Grok in vehicles. A software update in July 2025 added xAI's chatbot to Tesla vehicles, providing conversation without control over vehicle functions.

The xAI project. In March 2026 Musk unveiled a joint effort pairing xAI's Grok as a high-level navigator with a Tesla-developed agent processing real-time screen content.

Terafab. On 21 March 2026 SpaceX announced a joint project with Tesla and xAI to build a vertically integrated semiconductor megafabrication facility, addressing the companies' combined requirement for custom silicon.

The compensation package. In November 2025 shareholders approved a package for Musk valued at up to approximately US$1 trillion in Tesla shares, contingent on milestones including market capitalisation of US$8.5 trillion, 20 million vehicles delivered, one million robotaxis in operation and one million Optimus robots delivered. It followed the Delaware Court of Chancery's voiding of his 2018 package in January 2024, a decision that prompted the company's reincorporation in Texas.

Products

Vehicles

As of April 2026 Tesla sells the Model 3, a compact executive saloon; the Model Y, a compact crossover and the best-selling vehicle in the world in 2023; the Cybertruck; and the Semi, a class 8 lorry delivered in small numbers since 2022.

The Roadster of 2008, the Model S and the Model X have been discontinued, the latter two from the second quarter of 2026.

A second-generation Roadster has been promised since 2017 and repeatedly delayed, and the Cybercab is intended for autonomous operation.

Tesla's vehicles share several characteristics: minimal physical controls with functions moved to a central touchscreen, over-the-air software updates that add features after purchase, a battery as a structural floor member, direct sales without dealers, and a charging network the manufacturer built.

Energy

Powerwall is a home battery, Powerpack and Megapack are utility-scale storage systems, and Solar Roof and solar panels complete the range.

The energy business, long regarded as an afterthought, has grown very rapidly and is now among the most significant parts of the company. Megapack deployments, measured in gigawatt hours, have grown at rates far exceeding the vehicle business, and utility-scale storage is among the fastest-growing categories in energy globally, because grids with substantial renewable generation require it.

Tesla acquired SolarCity in 2016 for approximately US$2.6 billion, a transaction in which Musk was chairman and a substantial shareholder of both companies and which produced litigation alleging that Tesla had rescued a failing company in which he had an interest. He was found not liable in Delaware in 2022. The solar business has since declined substantially.

Software and autonomy

Autopilot provides lane keeping, adaptive cruise control and automatic lane changes and is included in all vehicles.

Full Self-Driving, sold as an option at prices between US$8,000 and US$15,000 over the years and now also by subscription, provides supervised autonomous driving on city streets. The name has been the subject of sustained criticism and of regulatory action in California, where a court found it misleading, because the system requires continuous supervision and is not self-driving under any technical definition.

Tesla's approach to autonomy differs from its competitors'. It uses cameras only, having removed radar in 2021 and ultrasonic sensors in 2022, and no lidar, on Musk's argument that human drivers use only vision and that lidar is a crutch. Every other significant autonomous programme, including Waymo's, uses lidar and radar in combination with cameras.

The approach's advantage is cost, which permits the hardware to be fitted to every vehicle sold, producing a fleet of millions collecting training data. Its disadvantage is that cameras fail in conditions where other sensors do not, and the National Highway Traffic Safety Administration has investigated Autopilot and Full Self-Driving repeatedly, including over collisions with stationary emergency vehicles and over operation in reduced visibility.

Dojo, a purpose-built training supercomputer announced in 2021, was reported to have been wound down in 2025 in favour of NVIDIA hardware and the AI5 inference chip.

Manufacturing

Tesla's factories are called Gigafactories, a name taken from the gigawatt hour scale of the Nevada cell plant and extended to all of them.

Fremont, California, the original vehicle plant, occupies the former NUMMI facility, a General Motors and Toyota joint venture that had closed in 2010 and that Tesla bought for approximately US$42 million. It produces Model 3, Model Y and, until 2026, Model S and Model X.

Gigafactory Nevada, from 2016, produces cells with Panasonic, battery packs, drive units and the Semi.

Gigafactory Shanghai, begun in January 2019 and producing within eleven months, is the largest and by most accounts the most efficient, supplying China and exporting to Europe and Asia. It was the first wholly foreign-owned car plant permitted in China.

Gigafactory Berlin-Brandenburg, from March 2022, supplies Europe, and its construction was delayed substantially by German environmental permitting and by local opposition concerning water abstraction in a region with limited supply.

Gigafactory Texas, from April 2022, is the headquarters and produces Model Y and Cybertruck.

Gigafactory New York at Buffalo produces Superchargers and previously solar products.

Tesla's manufacturing approach has several distinctive elements. Giga casting replaces dozens of stamped and welded parts with a single aluminium casting produced on very large machines from IDRA, which reduces cost, weight and assembly time, and which makes repair after a collision substantially more expensive. Structural batteries make the pack a load-bearing member. Vertical integration extends further than any competitor's, with Tesla making its own seats, its own motors, much of its own software and, through the Terafab project, potentially its own chips.

The unboxed process, announced for the next-generation vehicle, proposes assembling large subassemblies in parallel and combining them at the end rather than moving a body along a line, which if it works would be the largest change to automotive assembly since the moving line itself.

Sales and service

Tesla sells directly to customers rather than through franchised dealers, which is conventional everywhere except the United States, where laws in a number of states prohibit a manufacturer from selling directly in order to protect dealers.

The company has litigated and lobbied state by state for two decades, with mixed results, and in several states including Texas until recently a customer must buy a vehicle out of state.

The direct model gives Tesla control of the price, which is why it can change prices overnight, control of the customer relationship and the data, and no dealer margin, and it removes the service revenue that dealers earn and that offsets thin new car margins for conventional manufacturers.

Service has been a persistent complaint, with waiting times, parts availability and repair costs criticised, and insurance premiums for Tesla vehicles have been notably high because of the repair costs that giga casting and the integrated structure produce.

Superchargers number in the tens of thousands of connectors worldwide and are the largest fast-charging network. Tesla opened the standard as the North American Charging Standard in 2022, and between May 2023 and February 2024 substantially every North American manufacturer announced adoption, which made Tesla's connector the American standard and created a durable revenue stream from charging other manufacturers' vehicles.

The abrupt dismissal of the entire Supercharger team in April 2024, approximately 500 people, was among the most criticised decisions of the period, and several were subsequently rehired.

Finances

Tesla's revenue grew from approximately US$7 billion in 2016 to US$31.5 billion in 2020, US$81.5 billion in 2022 and approximately US$97 billion in 2023, and has declined modestly since as deliveries have fallen and prices have been cut.

It became consistently profitable from 2020, having lost money for most of its previous existence, and its automotive gross margin reached approximately 30 per cent in 2022, an extraordinary figure for a car manufacturer, before falling to roughly half that following the price reductions.

Regulatory credits have been a substantial profit source. Manufacturers that do not meet zero emission requirements buy credits from those that exceed them, and Tesla, which sells only electric vehicles, has sold billions of dollars of credits to competitors, including Stellantis, Honda and others. The revenue is close to pure profit and declines as competitors electrify.

The energy business has grown to a substantial proportion of revenue with better margins than vehicles.

Capital has been raised repeatedly through share issuance, and the company holds a very large cash balance. It bought US$1.5 billion of bitcoin in February 2021, sold most of it, and retains a holding.

The valuation has for a decade exceeded that of the rest of the automotive industry combined at several points, and is not explicable by automotive earnings. It rests on the expectation that autonomy, energy and robotics will produce returns of a different kind, and the argument between those who accept and those who reject that expectation is the whole of the discussion about the company's worth.

Elon Musk and Tesla

Tesla is more completely identified with one person than any company of comparable size, and the identification cuts both ways.

Musk's contributions are substantial and specific. He funded the company when nobody else would, took over when it was failing, made the judgements about the Model S, the Supercharger network and the Shanghai plant that determined its success, and sustained the manufacturing effort during 2017 and 2018 when institutional opinion expected failure.

His costs are equally specific and have grown. The funding secured episode produced a securities fraud settlement. The 2018 accusation that a British cave diver involved in the Thai rescue was a paedophile produced a defamation trial that Musk won. The 2022 Twitter acquisition consumed his attention and approximately US$44 billion, and required him to sell a very large quantity of Tesla shares. His political activity since 2024, including a substantial financial and personal role in the Trump campaign and the Department of Government Efficiency, has produced measurable commercial damage to Tesla in Europe and among American buyers, a group whose politics have historically been the opposite of his.

The board's independence has been questioned persistently. The 2018 compensation package was voided by the Delaware Court of Chancery in January 2024 on the ground that the process had not been at arm's length and that the disclosure was inadequate, which prompted reincorporation in Texas and a re-approval by shareholders. The 2025 package, at up to approximately US$1 trillion, was approved by a very large majority of voting shareholders, which its supporters describe as democracy and its critics as the consequence of a retail shareholder base that regards Musk as the company.

His attention is divided among Tesla, SpaceX, xAI, X, Neuralink and The Boring Company, and he has been explicit that the division is deliberate.

Criticism

Autopilot safety. The National Highway Traffic Safety Administration has conducted multiple investigations, has required recalls delivered by software update, and has examined collisions with stationary emergency vehicles, with motorcyclists and in reduced visibility. Tesla publishes safety statistics comparing Autopilot-engaged miles to the American average, which researchers have criticised as not comparing like with like, since motorway driving is safer than average regardless of the system.

Full Self-Driving naming has been found misleading by California's Department of Motor Vehicles and by regulators elsewhere.

Labour. The Fremont plant has been the subject of a long unionisation effort by the United Auto Workers. The National Labor Relations Board found in 2019 that Tesla had violated labour law, including by Musk's tweet threatening the loss of stock options, and ordered him to delete it. Racial harassment claims at Fremont produced a US$137 million jury award in 2021, later reduced, a California Civil Rights Department lawsuit alleging systemic discrimination, and a class action. Injury rates at Fremont were reported above industry averages, which Tesla disputed.

Working conditions in the Berlin plant produced a strike vote and complaints about sick leave management, and the Swedish mechanics' strike from October 2023, over Tesla's refusal to sign a collective agreement, produced sympathy action across Nordic industries including dockworkers, postal workers and cleaners that continued for more than a year, and is the most sustained European labour dispute the company has faced.

Quality. Panel gaps, paint and assembly defects have been documented consistently and have improved, and Tesla has conducted a very large number of recalls, many delivered by software.

Environmental permitting at Berlin, at Austin and at Fremont has produced repeated violations and fines.

Supply chain. Cobalt sourcing and the conditions in Democratic Republic of the Congo mining have been examined, and Tesla has moved substantially toward lithium iron phosphate chemistry containing no cobalt.

Musk's conduct is described above and is now the criticism most frequently made by the company's own customers.

Vehicles in detail

Roadster (2008 to 2012)

The first Tesla was built on a Lotus Elise glider, with the body extended and the chassis modified more substantially than the arrangement suggested, and Lotus supplied approximately 2,500 units.

Its battery contained 6,831 cylindrical lithium-ion cells of the 18650 format used in laptops, arranged in eleven sheets, with liquid cooling that was among the design's more significant contributions since thermal management is what allows lithium-ion cells to survive automotive duty cycles.

It accelerated to 60 miles per hour in 3.7 seconds in the Sport version and had an EPA range of 244 miles, which no electric vehicle had approached.

Its development problems were extensive. The original two-speed transmission could not survive the motor's torque and was eventually replaced with a single ratio, requiring a motor redesign. The cost per unit exceeded the selling price for a substantial period. Approximately 2,450 were sold in more than 30 countries.

One Roadster, Musk's personal car, was launched into a heliocentric orbit aboard the first SpaceX Falcon Heavy in February 2018.

Model S (2012 to 2026)

The Model S was designed from the ground up by Franz von Holzhausen, who had joined from Mazda, and it is the vehicle that established that an electric car could be better than a petrol one rather than merely acceptable.

Its architecture placed the battery as a flat slab in the floor, which lowered the centre of gravity below that of any production saloon, permitted a flat interior floor, and provided both a boot and a front compartment.

The 17-inch portrait touchscreen removed almost every physical control, which was contentious and which every manufacturer subsequently imitated to varying degrees.

Over-the-air updates were introduced with it, and the company has since added features including improved acceleration, new interface elements and Autopilot to vehicles already sold.

The Plaid version, from 2021, produces over 1,000 horsepower from three motors and accelerates to 60 miles per hour in under two seconds, which made it among the fastest accelerating production cars ever made.

Its discontinuation was announced for the second quarter of 2026.

Model X (2015 to 2026)

The Model X shares the Model S platform and adds falcon-wing rear doors, which open upward and which contain sensors to avoid obstacles in confined spaces.

The doors were the principal cause of a production delay of approximately two years and of persistent reliability problems, and Musk has described the vehicle as having been over-engineered and the doors as a mistake that nearly destroyed the company.

It was nonetheless the first electric sport utility vehicle of any capability and it established the category.

Model 3 (2017 to present)

The Model 3 is the vehicle that made Tesla a volume manufacturer.

Its interior removed almost every element, including the instrument cluster in front of the driver, leaving a single central screen, which was a cost decision presented as a design one and which divided opinion.

Its production difficulties are described above and it eventually reached output exceeding 5,000 a week.

The Highland update of 2023 revised the exterior, the interior and the suspension, and removed the indicator stalk, which was widely criticised and partly reversed.

Model Y (2020 to present)

The Model Y shares approximately three-quarters of its components with the Model 3 in crossover form, which is the segment that dominates global vehicle sales.

It became the best-selling vehicle of any kind in the world in 2023, the first electric vehicle and the first non-petrol vehicle to achieve that, and it remains Tesla's principal product.

The Juniper update of 2025 revised it comparably to the Model 3's Highland.

Cybertruck (2023 to present)

The Cybertruck was revealed in November 2019, at which event the demonstration of its armoured glass failed on stage when a steel ball broke a window, an incident Musk has described as a marketing benefit.

Its body is made of cold-rolled stainless steel that is not painted and that cannot be stamped into curves, which is why the vehicle is composed of flat planes. The construction is structurally unusual, with the exoskeleton bearing load rather than a separate frame.

Deliveries began in November 2023, three years later than announced and at prices substantially above those quoted. Sales have been far below the approximately one million reservations taken at US$100 each, recalls have been frequent, and its reception outside the United States has been limited by its dimensions and by regulations concerning pedestrian impact that its rigid edges do not satisfy.

Semi (2022 to present)

The Semi is a class 8 tractor unit with a claimed range of 500 miles, delivered in small numbers to PepsiCo and others since December 2022, with volume production repeatedly delayed and a dedicated plant in Nevada under construction.

Announced

The second-generation Roadster was announced in 2017 with claims including acceleration to 60 miles per hour in 1.9 seconds and, subsequently, rocket thrusters, and has not appeared.

The Cybercab, revealed in October 2024, is a two-seat vehicle without steering wheel or pedals intended for autonomous operation.

Batteries

Tesla's battery strategy has been among its most consequential technical choices.

Cell formats

18650 cells, 18 millimetres by 65, were the laptop standard and were used in the Roadster and Model S. The decision to use thousands of small commodity cells rather than a few large purpose-built ones gave Tesla access to a mature supply chain and to falling prices, and required sophisticated management to handle the thermal and failure behaviour of so many cells.

2170 cells, larger, were introduced with the Model 3 and are made with Panasonic at Gigafactory Nevada.

4680 cells, announced at Battery Day in September 2020, are substantially larger with a tabless design intended to reduce internal resistance and to permit faster charging and cheaper manufacture. Production has been difficult, particularly the dry electrode process, and has taken several years longer than announced.

Lithium iron phosphate cells, bought from CATL and BYD and used in standard range vehicles, contain no cobalt or nickel, are cheaper, tolerate full charging without degradation and have lower energy density. Their adoption across the industry has been one of the significant developments of the past five years and has substantially reduced concerns about cobalt sourcing.

Thermal and structural

Tesla's liquid cooling arrangement, in which a ribbon of cooling tube snakes between cells, was among the Roadster's genuine innovations and is why its batteries survived.

Structural packs, introduced with the 4680 vehicles, bond the cells into a unit that bears chassis loads, which removes mass and manufacturing steps and which makes the pack effectively unrepairable.

Supply

Tesla buys cells from Panasonic, LG and CATL, and manufactures its own at Nevada, Texas and Berlin. It has invested in lithium refining in Texas, in cathode production, and has taken positions in mining supply agreements, on the judgement that raw material availability rather than cell manufacturing is the eventual constraint.

Charging

The Supercharger network was a decision no other manufacturer would have made and it is among the principal reasons Tesla succeeded.

The proposition in 2012 was that an electric car is not usable for long journeys without fast charging, that no such network existed, that the public sector would not build one quickly and that no other manufacturer would build one for a competitor's customers. Tesla therefore built its own, initially free to use for Model S owners.

The network now comprises tens of thousands of connectors worldwide, is the largest and by most measures the most reliable fast-charging network, and its reliability is the specific advantage: the persistent complaint about competing networks is that a substantial proportion of chargers do not work.

The connector was opened as the North American Charging Standard in November 2022, and between May 2023 and February 2024 essentially every manufacturer selling in North America announced adoption, which made it the continental standard and gave Tesla a durable revenue stream from charging vehicles it did not make.

In Europe Tesla uses the CCS connector and has opened much of the network to other vehicles.

The dismissal of the entire Supercharger organisation in April 2024, approximately 500 people including its widely respected leader Rebecca Tinucci, halted expansion and was reversed in part within weeks, and is generally cited as the clearest example of the costs of the company's management style.

Autonomy in detail

Tesla's approach to autonomous driving is the most contested technical position in the automotive industry and the one on which a substantial part of its valuation rests.

The vision-only argument

Musk's position, stated consistently since approximately 2019, is that a human drives using two cameras and a neural network, that roads are designed for vision, and that a sufficiently capable vision system with a sufficiently good network will therefore drive. He has described lidar as a crutch and expensive appendix.

Tesla removed radar from new vehicles in 2021 and ultrasonic parking sensors in 2022, leaving eight cameras and the inference computer.

The counter-argument, made by essentially every other autonomous programme, is that redundancy matters more than sufficiency: cameras fail in fog, in direct sun, in heavy rain and at night in ways that radar and lidar do not, and an autonomous system without redundant sensing has no fallback when its primary sensor is degraded. Waymo, Cruise, Baidu, Mobileye and the Chinese manufacturers all use lidar.

The economic argument favours Tesla. Lidar cost tens of thousands of dollars a unit when the debate began, which made fitting it to every vehicle sold impossible, and cameras cost a few dollars, which permits a fleet of millions collecting data continuously. Lidar has since fallen to a few hundred dollars, which weakens the argument considerably, and the Chinese manufacturers now fit it as standard.

The data argument

Tesla's claimed advantage is fleet data: millions of vehicles driving billions of miles, from which the difficult situations can be mined and used for training.

The counter-argument is that data quantity is not the constraint and that what matters is the rare events, that labelling remains expensive, and that Waymo's approach of operating a small fleet in restricted areas with intensive supervision has produced a system driving without a human in the vehicle while Tesla's has not.

Progress

Autopilot from 2014 provides lane keeping and adaptive cruise control, which is Level 2 driver assistance and is comparable to systems from several manufacturers.

Full Self-Driving (Supervised), released in limited beta from October 2020 and to all American buyers who paid for it from 2022, attempts city driving with continuous human supervision. Its capability has improved substantially, particularly after the version 12 rewrite of 2024, which replaced approximately 300,000 lines of hand-written control code with an end-to-end neural network trained on video.

The end-to-end approach is a genuine technical shift: rather than perceiving objects, predicting their behaviour and planning a path in separate hand-engineered modules, a single network maps camera input to control output, learned from human driving examples. It produces markedly more natural behaviour and makes failures harder to diagnose, because there is no module to inspect.

Robotaxi service began in Austin in June 2025 with safety monitors in the front passenger seat, extended to a small number of other cities, and operates within geofenced areas with remote supervision, which is substantially the arrangement Waymo used in 2018.

The predictions. Musk has forecast full autonomy in the following year in 2016, 2017, 2018, 2019, 2020, 2021, 2022 and subsequently. In 2019 he stated that there would be one million robotaxis on the road in 2020. The consistency of the error is itself a datum, and it is the principal reason analysts discount the company's statements about the timeline while accepting that the capability is improving.

Regulation

The National Highway Traffic Safety Administration has opened several investigations, has required recalls delivered by software update including one covering approximately two million vehicles in December 2023 over driver monitoring, and has examined collisions with stationary emergency vehicles and operation in reduced visibility.

California's Department of Motor Vehicles found the Full Self-Driving name misleading, and a judge ruled in 2025 that the company had engaged in false advertising in its use of the terms.

Several jurisdictions including Germany have restricted the use of the names, and China requires specific approval for driver assistance features which Tesla obtained progressively.

Optimus

Optimus is the humanoid robot Tesla announced at its AI Day in August 2021, at which the concept was presented by a dancer in a costume, an event that produced considerable ridicule and that was followed by genuine prototypes from 2022.

The argument for a car company building a robot is that the required components are the ones a car company makes: actuators derived from motor expertise, batteries, computer vision from the driving programme, and an inference computer.

The argument against is that a bipedal robot is a substantially harder control problem than a car, that the applications are unproven, that Boston Dynamics has spent thirty years and has not produced a commercial product, and that the manufacturing volumes required to make it cheap do not exist until the product is useful.

Progress has been visible. The Gen 2 and Gen 3 prototypes walk, manipulate objects, and have hands with increasing degrees of freedom, and Tesla has said that units operate in its own factories performing limited tasks.

Musk has stated that Optimus will eventually be the most valuable product the company makes and has forecast production in the millions, and the decision announced on 28 January 2026 to discontinue the Model S and Model X to free capacity for it is the strongest indication that the company's priorities have shifted.

Demonstrations have been criticised. A widely covered event in October 2024 featured robots serving drinks and conversing, and it emerged that they were teleoperated by humans, which Tesla had not stated and which damaged the credibility of the programme.

Energy in detail

Tesla Energy has grown from a side business into a substantial one and its economics are better than the vehicle business's.

Megapack is a utility-scale battery container of approximately 3.9 megawatt hours, installed in arrays that store energy from renewable generation and discharge it when required. Deployments have grown at rates exceeding 50 per cent annually and are measured in tens of gigawatt hours.

The market exists because grids with substantial wind and solar generation require storage to manage the mismatch between generation and demand, and because batteries have become cheaper than the gas peaking plants that previously performed the function. The Hornsdale installation in South Australia, completed in 2017 in 100 days following a public wager, demonstrated that a large battery could stabilise a grid and earn money doing it, and it is generally credited with creating the market.

Powerwall is the residential equivalent, and the virtual power plant programmes, in which thousands of Powerwalls are dispatched collectively by a utility, are an early example of distributed energy resources operating as a grid asset.

Megafactory Lathrop in California and a second facility in Shanghai produce Megapacks, the latter opened in 2025 and supplying Asian markets.

Solar has declined substantially. The SolarCity acquisition brought a door-to-door sales operation and a leasing model that Tesla discontinued, the Solar Roof product has been produced in very small quantities against repeated announcements, and the business is now a small fraction of what SolarCity was.

Financial history

Tesla's annual revenue, approximately: US$204 million in 2011, US$2.0 billion in 2013, US$7.0 billion in 2016, US$21.5 billion in 2018, US$24.6 billion in 2019, US$31.5 billion in 2020, US$53.8 billion in 2021, US$81.5 billion in 2022, US$96.8 billion in 2023, and approximately US$98 billion in 2024 with a decline following.

It reported losses every year from its founding until 2020, cumulatively several billion dollars, and has been profitable since.

Net income reached approximately US$12.6 billion in 2022 and approximately US$15 billion in 2023, the latter including a one-off tax benefit of approximately US$5.9 billion from recognising deferred tax assets.

The share price, adjusted for the five-for-one split in August 2020 and the three-for-one split in August 2022, rose from an equivalent of approximately US$1.13 at listing to above US$400 at the 2021 peak, fell by approximately 70 per cent through 2022, recovered, and has been volatile since.

The company has raised approximately US$20 billion through share issuance over its life, has repaid its government loans, and holds a cash balance exceeding US$30 billion.

Regulatory credit revenue has totalled several billion dollars and exceeded net income in several years, meaning that Tesla's profitability in those years depended entirely on payments from competitors who had not electrified.

Software features

Tesla vehicles run the company's own operating system on its own computers, and the resulting features are a substantial part of what customers describe as differentiating the product.

Over-the-air updates arrive several times a year and have added or improved acceleration, range, braking distance, interface elements, driver assistance, entertainment and diagnostics. A vehicle bought in 2018 has received functionality that did not exist when it was sold, which is a fundamental change to the relationship between manufacturer and owner and which every competitor has attempted to replicate with varying success.

Sentry Mode records the vehicle's surroundings when parked and has produced a very large quantity of evidence in criminal proceedings, which has in turn raised questions about a manufacturer operating a distributed surveillance network.

Dog Mode, Camp Mode and Bioweapon Defense Mode, the last a HEPA filtration system named with characteristic understatement, address specific uses.

Entertainment includes streaming video, games, karaoke and, since July 2025, the xAI chatbot Grok, which converses but does not control vehicle functions.

Navigation integrates the Supercharger network, routing automatically to chargers, preconditioning the battery before arrival to permit faster charging, and estimating arrival state of charge, which is the specific feature that makes long journeys practical and which competitors have been slow to match.

Smart Summon moves the vehicle to the owner in a car park without a driver, a feature whose regulatory status has been contested and whose reliability has been criticised.

The mobile application controls climate, locking, charging and location and serves as the key, and the physical key card exists principally as a backup.

Diagnostics are conducted remotely, and a substantial proportion of service issues are resolved without the vehicle being seen, which is the source both of Tesla's low service cost and of the complaint that it is difficult to obtain a physical appointment.

Criticism of the software approach

The removal of physical controls has been criticised by safety researchers and by Euro NCAP, which from 2026 requires physical controls for certain functions in order to achieve its highest rating, on evidence that touchscreen operation takes the driver's eyes from the road for longer.

The removal of the indicator stalk in the 2023 Model 3 update, replacing it with buttons on the steering wheel, was widely criticised and partly reversed.

Feature paywalls, in which hardware already fitted to a vehicle is enabled by payment, have produced objection, most notably where acceleration was sold as a software unlock and where heated rear seats required activation.

Musk's other companies

Tesla's relationship with Musk's other ventures is closer than is usual and has been the subject of governance criticism.

SpaceX shares no formal relationship and has shared personnel, manufacturing knowledge and, during Tesla's 2008 crisis, Musk's personal capital allocated between them. The Terafab semiconductor project announced on 21 March 2026 is a formal joint undertaking of SpaceX, Tesla and xAI.

xAI provides the Grok model integrated into Tesla vehicles from July 2025, and the March 2026 joint project pairs Grok with a Tesla agent. Tesla shareholders voted in 2025 on whether the company should invest in xAI, and the proposal was advisory. Musk has said that Tesla's artificial intelligence work and xAI's are complementary rather than competing, which several shareholder groups have disputed on the ground that Tesla's own research talent has moved to xAI.

X is where Musk conducts the communication that has affected Tesla's brand, and the 2022 acquisition required him to sell approximately US$23 billion of Tesla shares, which contributed to the share price decline of that year.

The Boring Company built the Las Vegas Convention Center Loop, in which Teslas are driven through tunnels, and has a commercial relationship with Tesla.

Neuralink has no operational relationship.

The general criticism is that a chief executive running six companies cannot give any of them adequate attention, that transactions between them are not arm's length, and that talent and technology move between them in ways shareholders cannot observe. Musk's answer is that the companies are complementary and that he allocates his time to whichever is in the greatest difficulty.

Analysis of the valuation

Tesla's market capitalisation has at several points exceeded the combined value of the next several manufacturers, and the argument about whether this is justified is the central question about the company.

The automotive case does not support it. Tesla sells approximately 1.6 million vehicles a year against Toyota's 10 million and Volkswagen's 9 million, its margins have compressed toward the industry's, and its growth has reversed. Valued as a car company on any conventional measure it is worth a fraction of its price.

The autonomy case is the principal argument for the valuation. If Tesla achieves unsupervised autonomy, its existing fleet of several million vehicles becomes a revenue-generating asset, the marginal cost of a robotaxi mile approaches the electricity and the depreciation, and the addressable market is the entire value of personal transport and of the taxi industry. The case depends on a capability that has been forecast annually for a decade and has not arrived, and on a regulatory environment that does not yet exist.

The energy case is the least discussed and the most concrete. Grid storage demand is growing very rapidly, the margins are better than automotive, and Tesla holds a strong position with a proven product. It is not large enough to justify the valuation alone.

The robotics case is the newest. If humanoid robots become a general product, the addressable market is the labour force, which is larger than any other market. Nothing about this is established, no working product exists, and Musk's assertion that it will be the company's most valuable business is an assertion.

The bear case is that the automotive business is a conventional manufacturer facing a stronger Chinese competitor, that the autonomy and robotics claims have a documented history of being wrong about the timing, and that the price embeds an expectation of a transformation that the company has been predicting for a decade.

Both positions have been argued for years by intelligent people and the disagreement has not narrowed, which is itself informative: the question is not about the facts, which both sides accept, but about a probability that cannot be estimated from them.

Notable people

Elon Musk has been chief executive since October 2008 and the company's defining figure since 2004.

Martin Eberhard co-founded the company, was chief executive until 2007, and has been publicly critical of Musk's account of the founding.

Marc Tarpenning co-founded it and left in 2008.

JB Straubel was chief technology officer from 2004 to 2019 and is regarded within the industry as having been responsible for the battery architecture that made the vehicles possible. He founded Redwood Materials, a battery recycling company, and returned to Tesla's board in 2023.

Franz von Holzhausen, chief designer since 2008, designed every Tesla production vehicle from the Model S onward.

Drew Baglino, senior vice president for powertrain and energy, left in April 2024 during the redundancy programme.

Andrej Karpathy directed artificial intelligence from 2017 to 2022 and was responsible for much of the vision architecture.

Zachary Kirkhorn was chief financial officer from 2019 to 2023, succeeded by Vaibhav Taneja.

Tom Zhu, who ran the Shanghai operation, was brought to oversee global production in 2023 and returned to Asia.

Robyn Denholm has chaired the board since 2018.

Selected statistics

• Vehicles delivered: approximately 1.64 million in 2025, down from a peak of approximately 1.81 million in 2023.
• Cumulative vehicles: more than 7 million.
• Superchargers: tens of thousands of connectors at several thousand stations worldwide.
• Energy storage deployed: measured in tens of gigawatt hours annually and growing rapidly.
• Employees: approximately 125,000.
• Gigafactories: six operating.
• Revenue: approximately US$98 billion at peak.
• Market capitalisation: exceeded US$1 trillion in several periods.
• Full Self-Driving price: between US$8,000 and US$15,000 over its history, and available by subscription.
• Musk's 2025 compensation package: up to approximately US$1 trillion in shares on milestones.

Gigafactories

Fremont

The Fremont plant occupies the former New United Motor Manufacturing Inc. facility, a joint venture between General Motors and Toyota established in 1984 to teach American workers and managers the Toyota Production System.

NUMMI was among the more consequential industrial experiments in American history: a plant that had been General Motors' worst, closed in 1982 with an absentee rate approaching 20 per cent, reopened under Toyota's system with substantially the same workforce and became one of its best.

It closed in 2010 when Toyota withdrew after General Motors' bankruptcy, and Tesla bought it for approximately US$42 million, a very small sum for a facility of that size and capability, together with equipment.

It now produces Model 3 and Model Y and is among the most productive car plants in North America by vehicles per employee, which is notable given the production difficulties of 2017 and 2018.

Nevada

Gigafactory Nevada, near Sparks, was announced in 2014 as a joint venture with Panasonic, which operates the cell production lines within a building Tesla owns.

The scale was the point. The facility was designed to produce more lithium-ion cells annually than the entire world output when it was announced, on the judgement that cell supply would constrain the industry and that no supplier would build the capacity speculatively.

It also produces battery packs, drive units and the Semi, and an expansion announced in 2023 committed several billion dollars to 4680 cell and Semi capacity.

Shanghai

Gigafactory Shanghai is the company's most efficient plant and the clearest demonstration of what Chinese industrial policy can deliver.

Construction began in January 2019 and vehicles were produced within eleven months, which is approximately a third of the time a comparable plant takes elsewhere. The Chinese government provided land, low-interest loans and, critically, permission to operate wholly owned, which no foreign car manufacturer had previously received.

The plant supplies China, Europe, Australia and much of Asia, produces at a cost substantially below Fremont's, and its supply chain is almost entirely Chinese, which is both the source of the cost advantage and a strategic exposure as trade restrictions tighten.

The Megafactory at Shanghai, producing Megapacks, opened in 2025.

Berlin

Gigafactory Berlin-Brandenburg at Grünheide opened in March 2022 after delays of approximately two years caused by German permitting, which requires environmental assessment, public consultation and provisional approvals at each construction stage.

Local opposition concerned water abstraction in a region of sandy soil and limited groundwater, and protests including a 2024 arson attack on an electricity pylon that halted production for several days have continued.

The plant produces Model Y for Europe and its output has been affected by the decline in European sales since 2024.

Texas

Gigafactory Texas at Austin opened in April 2022 and is the corporate headquarters. It produces Model Y and Cybertruck and houses 4680 cell production, and its scale, approximately 10 million square feet of floor area, makes it among the largest buildings in the world by volume.

Other

Buffalo, New York, produces Superchargers and previously produced solar products under an arrangement with the state that required specified employment levels and that has been a subject of political controversy.

Lathrop, California, produces Megapacks.

Announced facilities have included Mexico, at Monterrey, approved in 2023 and paused, and India, repeatedly discussed.

Manufacturing methods

Tesla's manufacturing has been both innovative and, at times, the source of its worst difficulties.

Giga casting replaces the rear and front underbody, previously approximately 70 stamped and welded parts, with a single aluminium casting produced on machines of 6,000 to 9,000 tonnes of clamping force supplied principally by IDRA of Italy. The benefits are fewer parts, less weight, less assembly labour and less factory floor area. The costs are that a collision damaging the casting may write off the vehicle, that repair is difficult, and that a defect in a casting scraps a large component.

Structural battery packs integrate the cells into a unit that bears chassis loads, removing the separate pack enclosure and the floor structure.

The unboxed process, announced in 2023 for the next-generation vehicle, proposes building large modules in parallel, painting and fitting them separately, and joining them at the end, which would eliminate the moving assembly line that Ford introduced in 1913. Whether it works is unknown, since the vehicle it was designed for has not appeared.

Automation. The 2017 attempt to automate final assembly extensively failed, and Musk's public acknowledgement that excessive automation was a mistake is among the more candid statements any chief executive has made about a strategic error. The company has since automated selectively.

Vertical integration extends to seats, motors, inverters, software, and increasingly chips, which is the opposite of the industry's thirty-year trend toward outsourcing and which Tesla justifies on the ground that a supplier optimising its own component does not optimise the vehicle.

Workforce and culture

Tesla employs approximately 125,000 people, a figure reduced by approximately 10 per cent in a redundancy programme announced in April 2024.

Its culture is described consistently by employees as extremely demanding, with long hours, rapid decisions, frequent reorganisation and very little process, and as offering work of a kind and at a pace unavailable elsewhere.

Unionisation has been resisted comprehensively. The United Auto Workers' campaign at Fremont from 2017 produced National Labor Relations Board findings against the company, including over Musk's 2018 tweet suggesting that employees would lose stock options if they unionised, which the board ordered deleted and which litigation over the remedy has continued.

The Swedish dispute beginning in October 2023, in which the mechanics' union IF Metall struck over Tesla's refusal to sign a collective agreement, produced sympathy action of a kind peculiar to the Nordic system: dockworkers refused to unload Teslas, postal workers refused to deliver registration plates, cleaners refused to clean the showrooms, and electricians refused to service the chargers. Tesla imported plates by other means and sued the postal service. The dispute has continued for years and is the most sustained challenge to the company's labour position anywhere.

Discrimination. A jury awarded Owen Diaz US$137 million in 2021 over racial harassment at Fremont, reduced on appeal and retried; the California Civil Rights Department sued alleging systemic discrimination; and a class action covering thousands of Black employees has proceeded.

Safety. Reporting has documented injury rates at Fremont above industry averages and allegations that injuries were misclassified to reduce reportable figures, which Tesla has disputed.

Environment

Tesla's environmental argument is that its products displace petrol vehicles and fossil generation and that the resulting emissions reduction exceeds the emissions of manufacture.

The argument is well supported. Life cycle analyses consistently find that an electric vehicle produces higher manufacturing emissions, principally from battery production, and lower use-phase emissions, and that the crossover occurs within one to two years of typical driving in most grids and sooner in clean ones.

The complications are real. Battery production requires lithium, nickel, cobalt, graphite and copper, whose extraction has substantial local environmental and social effects, particularly cobalt in the Democratic Republic of the Congo and lithium in the Atacama. Tesla's shift toward lithium iron phosphate removes cobalt and nickel from a growing proportion of its vehicles.

Tesla's own operations have produced repeated environmental violations. The Fremont plant has been cited by the Bay Area Air Quality Management District on many occasions for emissions from its paint shop and foundry, and settled for approximately US$1.5 million in 2022; the Environmental Protection Agency reached a settlement over Clean Air Act violations in 2024; and the Berlin plant's water use and the Austin plant's wastewater have each produced regulatory findings.

The company publishes an impact report and does not publish a full corporate carbon footprint in the form its competitors do, and it has declined to participate in several standard disclosure frameworks.

Suppliers

Tesla's supply chain is less extensive than a conventional manufacturer's because of vertical integration, and its relationships with the suppliers it does use have been characterised as demanding.

Cells come from Panasonic, LG Energy Solution and CATL, and increasingly from Tesla's own lines.

Semiconductors come from Samsung, which fabricates the AI4 and AI5 inference chips Tesla designs, from TSMC, and from the conventional automotive suppliers for the several hundred other chips a vehicle requires. The semiconductor shortage of 2021, which halted production across the industry, affected Tesla substantially less because its software team rewrote firmware to use whatever chips were available, a flexibility conventional manufacturers did not have because their suppliers wrote the firmware.

Castings machines from IDRA, motors made internally, glass from a Tesla operation and external suppliers, and a long tail of conventional automotive suppliers complete the chain.

The company has been reported to demand price reductions retrospectively and to change specifications at short notice, practices that are not unusual in the industry and that a company with Tesla's growth rate can impose more effectively than most.

Policy and incentives

Tesla's growth has been assisted by public policy to an extent its founder's libertarian statements do not acknowledge, and the company has also been among the more effective lobbyists against policies that would assist its competitors.

Direct support has included the US$465 million Department of Energy loan of 2010, repaid early; Californian and Nevadan state incentives worth well over a billion dollars for the Nevada and Fremont facilities; New York State's approximately US$750 million investment in the Buffalo plant; and Chinese land, finance and regulatory accommodation at Shanghai.

Consumer incentives including the American federal tax credit, at US$7,500 per vehicle in various forms, and equivalents in Norway, the Netherlands, Germany, China and elsewhere have been central to electric vehicle adoption, and Tesla's sales have fallen measurably when they have been withdrawn, as in Germany in December 2023.

Regulatory credits have produced several billion dollars of revenue from competitors required to meet zero emission standards, and this is the mechanism by which the California Air Resources Board's rules, first adopted in 1990, transferred money from established manufacturers to Tesla for approximately fifteen years.

Musk's subsequent position has been to advocate the removal of the consumer credit, on the reasoning that Tesla would be less damaged by its removal than its competitors would, which is probably correct and which represents a change from the company's earlier advocacy.

Trade policy. Tariffs on Chinese vehicles in the United States, at 100 per cent from 2024, and in the European Union, at rates up to approximately 35 per cent from late 2024, protect Tesla's Western production from BYD and the other Chinese manufacturers, while Tesla itself exports from Shanghai to Europe and is subject to the European tariff at a lower rate negotiated individually.

Safety and recalls

Tesla vehicles have performed well in crash testing, achieving five-star ratings from the National Highway Traffic Safety Administration and Euro NCAP, assisted by the absence of an engine, which permits a larger crumple zone, and by the low centre of gravity, which reduces rollover.

Battery fires occur and are extensively reported. The available data indicate that electric vehicles catch fire less frequently than petrol ones per mile driven, and that when they do the fire is harder to extinguish because a thermal runaway in a lithium-ion pack sustains itself without external oxygen and can reignite.

Recalls have been numerous and the great majority have been delivered by software update, which is a genuine advantage over a physical recall and which has produced the statistical artefact that Tesla appears to issue more recalls than manufacturers whose defects require a workshop visit and are therefore addressed more slowly.

Significant recalls have included approximately two million vehicles in December 2023 over Autopilot driver monitoring; a recall over the Full Self-Driving beta's behaviour at intersections; Cybertruck recalls over the accelerator pedal, the windscreen wiper, the trim and other items; and a very large Chinese recall in 2023 covering driver assistance.

Investigations by the National Highway Traffic Safety Administration have concerned collisions with stationary emergency vehicles, phantom braking, steering wheel detachment, door handles and the operation of Full Self-Driving in low visibility.

Reception

Tesla's reception has been more polarised than that of almost any company and the polarity has reversed.

Among early customers the vehicles produced an enthusiasm unusual for a car, and the owner community functioned as an unpaid marketing organisation for a decade.

Among automotive journalists the Model S was greeted as a genuine advance and the subsequent products as competent, with persistent criticism of build quality, of service and of the removal of physical controls.

Among engineers the company is respected for the battery architecture, the Supercharger network, the Shanghai plant and the software, and criticised for the autonomy claims.

Among the established industry Tesla was dismissed for a decade, taken seriously from approximately 2018, and is now studied closely, with Volkswagen's and Toyota's senior executives having said publicly that Tesla's cost structure and software capability exceed theirs.

Among investors the company has been the most contested large stock of the period, with the short interest at times among the highest of any large company and with a retail following of unusual intensity.

Among the general public the position has changed. Surveys in the United States and Europe since 2024 have found Tesla's favourability falling sharply, particularly among the politically progressive urban buyers who were its core market, and rising modestly among conservatives who are less likely to buy an electric vehicle. The net effect has been negative in every market where it has been measured.

Comparison with the established industry

The differences between Tesla and a conventional manufacturer are structural and are worth setting out.

Vertical integration. A conventional manufacturer buys approximately 70 to 80 per cent of a vehicle's value from suppliers; Tesla makes substantially more itself, which gives it control of cost and of development pace and exposes it to the difficulty of doing many things.

Software. Tesla writes its own vehicle operating system, its own infotainment, its own driver assistance and much of its own manufacturing software. A conventional manufacturer integrates software from dozens of suppliers, each controlling its own module, which is why Volkswagen's software subsidiary Cariad has cost several billion euros and delayed multiple models.

Distribution. Direct sales rather than dealers.

Model cadence. Conventional manufacturers replace a model every six to seven years with a facelift halfway; Tesla updates continuously by software and infrequently by hardware, which reduces engineering cost and produces a range that ages visibly.

Margins. Tesla's automotive gross margins at their peak were approximately double the industry's, which reflected the absence of dealer margin, the vertical integration, the limited model range and, substantially, a price point buyers accepted before competition existed.

Labour. Almost entirely non-union in the United States, against a fully unionised Detroit industry whose 2023 agreement raised costs substantially.

Capital. Tesla built five plants from nothing in a decade; the established industry has been closing plants.

The electric vehicle market

Tesla's position is best understood against the market it created.

Global electric vehicle sales grew from a few thousand in 2010 to approximately 17 million in 2024, of which the majority are in China, and the growth rate in Europe and North America slowed substantially from 2023.

The slowdown has several causes: incentive withdrawal, charging infrastructure that remains inadequate outside a few countries, high interest rates affecting a purchase that costs more upfront and less to run, and the exhaustion of the early adopter population.

China is the exception. Electric and plug-in hybrid vehicles exceeded half of new sales in 2024, driven by BYD's pricing, by domestic competition of extraordinary intensity, and by policy sustained over fifteen years.

Hybrids have recovered substantially in Western markets as consumers hesitate, which Toyota had predicted and had been criticised for, and which has been the principal commercial vindication of its strategy.

The competitive question for Tesla is whether its advantages, cost, software, charging and brand, survive a market in which BYD sells comparable vehicles for substantially less and every manufacturer sells something electric.

Summary

Tesla built the electric car industry and is no longer the largest participant in it.

The company's achievement is that the world's automotive manufacturers are electrifying, which they were not doing and did not intend to do, and that they are doing so using an architecture, a charging standard and a set of manufacturing methods that Tesla established. Very few companies change an industry of that size and none has done so from a standing start within twenty years.

The company's present condition is that its vehicle range is ageing, its volume has been overtaken, its margins have compressed, its sales have fallen for two consecutive years, and its valuation depends on products, robotaxis and humanoid robots, that have been promised for longer than they have existed.

The determining variable is no longer technical. Tesla's engineering remains good and its cost position remains defensible outside China. What has changed is that its chief executive has become, for a substantial proportion of the people most likely to buy its products, a reason not to, and that is a problem for which the company has no engineering solution and about which its board has shown no inclination to act.

Timeline

• 2003 Incorporated on 1 July by Martin Eberhard and Marc Tarpenning.
• 2004 Elon Musk leads the Series A with US$6.5 million and becomes chairman; JB Straubel joins.
• 2006 The Roadster is revealed; the Secret Master Plan is published.
• 2007 Eberhard is removed as chief executive.
• 2008 Roadster deliveries begin; Musk becomes chief executive in October with the company near insolvency.
• 2009 Daimler invests US$50 million; the founder dispute is settled.
• 2010 A US$465 million Department of Energy loan is approved; Tesla lists on Nasdaq on 29 June; the NUMMI plant is acquired.
• 2012 The Model S is delivered from June; the first Superchargers open.
• 2013 Motor Trend names the Model S Car of the Year; the government loan is repaid nine years early.
• 2014 The patent pledge opens Tesla's patents; Autopilot hardware is introduced; Gigafactory Nevada is announced.
• 2015 The Model X is delivered; Autopilot is activated by software update; Powerwall is announced.
• 2016 The Model 3 is revealed in March with approximately 400,000 reservations; SolarCity is acquired.
• 2017 Model 3 production begins and proceeds very badly.
• 2018 The 5,000 per week target is reached in June; the funding secured episode in August produces the SEC settlement in September.
• 2019 Gigafactory Shanghai is built in eleven months; the Cybertruck is revealed in November.
• 2020 The Model Y is delivered; Tesla becomes consistently profitable; it joins the S&P 500 in December; the shares rise approximately eightfold.
• 2021 Gigafactory Berlin and Texas open; the headquarters moves to Austin; Optimus is announced in August; the market capitalisation exceeds US$1 trillion.
• 2022 Musk acquires Twitter in October; the North American Charging Standard is opened; the Semi is delivered in December.
• 2023 The Cybertruck is delivered in November; the Model Y becomes the world's best-selling vehicle; competitors adopt the charging standard.
• 2024 The Delaware court voids the 2018 compensation package in January; Tesla reincorporates in Texas; deliveries fall for the first time; the Supercharger team is dismissed in April; the Cybercab is revealed in October.
• 2025 European sales fall sharply amid political controversy; robotaxi service begins in Austin in June with safety monitors; Grok is added to vehicles in July; the approximately US$1 trillion compensation package is approved in November; UK registrations fall more than 29 per cent in December.
• 2026 BYD overtakes Tesla in battery electric deliveries in January; the Model S and Model X are announced for discontinuation on 28 January; the xAI joint project is unveiled in March; Terafab is announced on 21 March.

The patent pledge

On 12 June 2014 Tesla published a statement titled "All Our Patent Are Belong To You", announcing that it would not initiate patent litigation against anyone using its technology in good faith.

The reasoning Musk gave was that Tesla's competition was not other electric vehicle manufacturers but the very large number of petrol vehicles produced annually, that the electric vehicle programmes of the established manufacturers were small and unambitious, and that the industry's slow progress rather than intellectual property theft was the threat.

The pledge's practical effect has been disputed. Its terms require the user not to assert patents against Tesla, not to challenge Tesla's patents and not to market products similar to Tesla's, which lawyers have noted is broader than a straightforward licence, and Toyota cited the conditions when declining to use it.

Its symbolic effect was considerable and it is cited in discussions of open innovation.

Insurance

Tesla Insurance, launched in California in 2019 and extended to several states, prices premiums using data from the vehicle itself, producing a safety score from driving behaviour and adjusting the monthly premium accordingly.

The argument is that Tesla knows more about how its vehicles are driven and how much they cost to repair than any insurer, and that conventional insurers therefore overcharge.

The counter-argument, which has emerged in practice, is that Tesla vehicles are expensive to repair because of the structural design and the limited independent repair network, that insurers have priced accordingly, and that Tesla's own loss experience has reflected the same costs.

The business has grown slowly and has faced the regulatory approval processes that insurance requires state by state.

Direct sales and the dealer laws

American franchise laws, enacted in most states between the 1930s and the 1970s, prohibit a manufacturer from selling vehicles directly to consumers, in order to protect independent dealers from being undercut by the manufacturers whose products they sell.

The laws were written for an industry in which every manufacturer used dealers, and they did not contemplate a manufacturer with none.

Tesla has litigated and lobbied state by state for two decades. It sells directly in most states, operates galleries where staff may not discuss price in some, and cannot sell at all in others including, for long periods, Texas, where customers ordered online from out of state and took delivery at the state line.

Dealer associations have been among the most effective lobbying organisations in American state politics, since a dealership is a substantial local employer and a significant political donor in every legislative district, and the persistence of the laws against a general consensus among economists that they raise prices is a standard example of concentrated interests defeating diffuse ones.

Tesla in culture

Tesla occupies a position in public discussion out of proportion to its size, and the reasons have changed over time.

For approximately a decade it was the company that demonstrated that electric vehicles could be desirable, and the Model S in particular functioned as a status object in a way no previous environmentally motivated product had.

Its customer base was for most of that period politically progressive, affluent and concentrated in coastal American cities and in northern Europe, which is precisely the demographic most alienated by Musk's conduct since 2022, and the resulting reversal is among the more striking examples of a founder damaging a brand.

The share price is followed by a very large retail investor community whose engagement resembles that of a sports following more than of a shareholder base, and which has been a significant factor in the valuation.

In media Tesla appears constantly, and the company has never paid for advertising in the conventional sense, relying on the attention its products and its founder generate, which is an approach no other manufacturer could adopt.

Protests against the company since 2025, including vandalism of vehicles and showrooms and the organised Tesla Takedown movement, are unusual in being directed at a manufacturer's customers as well as at the company.

Corporate governance

Tesla's board has been criticised persistently for insufficient independence from Musk.

Its members have included his brother Kimbal Musk, Antonio Gracias, a long-standing business associate, and James Murdoch, and the Delaware Court of Chancery's January 2024 judgment voiding the 2018 compensation package found specifically that the process had not been at arm's length, that the negotiation had been conducted by people with close ties to Musk, and that shareholders had not been adequately informed.

Robyn Denholm became chair in 2018 under the SEC settlement, and her compensation, which was very large, was itself the subject of litigation settled in 2024 with a substantial repayment.

The company reincorporated in Texas in June 2024 following a shareholder vote, moving from Delaware, whose courts have the most developed corporate law in the United States and whose judgment had gone against Musk, to a jurisdiction with a newly created business court and far less precedent.

The 2025 compensation package, approved by approximately three-quarters of voting shares, is structured in tranches vesting on milestones including market capitalisation of US$8.5 trillion, adjusted earnings targets, 20 million cumulative vehicle deliveries, one million robotaxis in commercial operation, one million Optimus units delivered and ten million active Full Self-Driving subscriptions.

Its supporters argue that it pays nothing unless extraordinary value is created. Its critics argue that the sums are unprecedented, that the milestones are partly within Musk's power to define, and that the package exists because the board concluded it could not risk his departure, which is itself a description of a governance failure.

Effect on the industry

Tesla's effect on the automotive industry is the largest any company has had on any industry in recent decades, and the specific changes are worth listing.

Electrification. Every large manufacturer now sells electric vehicles and has committed to further electrification, and none did before Tesla demonstrated that people would buy them. The established industry's position until approximately 2015 was that electric vehicles were a regulatory obligation to be met at minimum cost.

The battery as architecture. Tesla established that an electric vehicle should be designed around a flat structural battery rather than adapted from a petrol platform, and every serious competitor now does the same.

Over-the-air updates. A car that improves after purchase was unknown before the Model S and is now general, and it has changed how manufacturers think about the product's life.

Direct sales. Tesla's refusal to use dealers has been followed by several manufacturers in Europe and has produced legislative conflict in the United States, and the agency model, in which the manufacturer sets the price and the dealer takes a commission, has spread substantially.

Charging. Building a charging network was regarded as a public sector problem until Tesla built one, and the adoption of its connector as the North American standard from 2023 makes its infrastructure the industry's.

Software. The interior with almost no physical controls, the central screen and the manufacturer's own operating system are now general, and the industry's difficulty in producing software of comparable quality has been among the clearest demonstrations of how different the competence required is from the one it had.

Valuation. The market's willingness to value a manufacturer at a multiple of a technology company rather than of an industrial one changed how the entire sector is analysed, and produced a wave of electric vehicle startups from 2020 of which the great majority have failed.

Manufacturing. Giga casting and structural batteries have been adopted by Toyota, Volvo, Hyundai and the Chinese manufacturers, and Toyota's decision to adopt casting after decades of refining its own production system is a substantial acknowledgement.

Competition

BYD is the most consequential competitor. A Chinese company that began making batteries and moved into vehicles, it manufactures its own cells using lithium iron phosphate chemistry, its own semiconductors and most of its own components, sells at prices Tesla cannot match, and overtook Tesla in battery electric deliveries in January 2026. Its expansion into Europe, South America and Southeast Asia is the principal competitive threat.

Chinese manufacturers including Geely, which owns Volvo, Polestar and Zeekr, SAIC, which owns MG, Nio, Xpeng, Li Auto and Xiaomi, whose SU7 sold in very large numbers on launch, collectively produce the majority of the world's electric vehicles and are the industry's technological pace-setters in several respects.

Established manufacturers including Volkswagen, Toyota, Hyundai, General Motors, Ford, Stellantis and Mercedes-Benz have each invested tens of billions of dollars, have generally lost money on electric vehicles, and have moderated their commitments since 2024 as demand growth slowed in Europe and the United States.

Rivian and Lucid are the surviving American startups, both loss-making and both dependent on external funding, the first from Volkswagen and the second from Saudi Arabia's Public Investment Fund.

In autonomy the competition is Waymo, owned by Alphabet, which operates driverless commercial service in several American cities with a sensor-heavy approach, and Chinese operators including Baidu's Apollo Go and Pony.ai.

In energy storage the competitors are BYD, CATL, Fluence, Samsung SDI and LG.

In humanoid robotics the competitors are Figure, Agility, Boston Dynamics under Hyundai, Unitree and a number of Chinese manufacturers, and the field is at a stage where nobody has a product.

Assessment

Tesla's achievement is settled and its future is not.

The achievement is that the automotive industry is electrifying and would not be without it. Every established manufacturer had the capability to build an electric car that people wanted and none did, because the business case did not exist until somebody demonstrated the demand. Tesla demonstrated it, at enormous cost and with several near-failures, and the industry followed. That is a change of a kind that very few companies produce.

The subsidiary achievements are substantial. The Supercharger network solved the problem that had made electric vehicles impractical. The Shanghai plant is among the best car factories in the world. The energy storage business is a genuine and growing position in a market that will be very large. And the company's manufacturing innovations have been adopted by Toyota, which is not an organisation that adopts other people's methods lightly.

The difficulties are also clear. Tesla's vehicle range is ageing, its price advantage has gone, BYD has overtaken it in volume, its European sales have fallen for reasons unrelated to its products, and the mass-market vehicle promised for a decade has not appeared. Deliveries have declined for two consecutive years.

The valuation depends entirely on the things that have not happened. Autonomy has been promised annually since 2016 and the robotaxi service operates with safety monitors in a small number of cities. Optimus has been demonstrated and not sold. The company that is worth more than the rest of the industry combined is worth that on the basis of products it does not yet have.

The final complication is the one Tesla cannot engineer around. Its chief executive's public conduct has measurably reduced demand among the people most likely to buy its vehicles, in the markets where electric vehicle adoption is highest, and no product decision addresses it.

See also

• Elon Musk
• BYD
• SpaceX
• Electric vehicle
• Self-driving car
• Waymo
• Lithium-ion battery

The master plans

Tesla has published four strategic documents, and comparing them to what happened is a useful exercise.

The Secret Master Plan, 2006

Musk published four sentences in August 2006: build a sports car; use that money to build an affordable car; use that money to build an even more affordable car; and while doing above, provide zero emission electricity generation options.

It was executed almost exactly. The Roadster funded the Model S, which funded the Model 3, and SolarCity provided the generation element, though that has since declined substantially.

The plan's accuracy over a decade is unusual and is the strongest evidence for taking the company's stated intentions seriously, which is the argument its supporters make about the current ones.

Master Plan Part Deux, 2016

The second plan proposed integrating energy generation and storage, expanding to cover the major vehicle segments including heavy lorries and buses, developing self-driving capability approximately ten times safer than manual driving, and enabling owners to add their vehicles to a shared fleet earning money while not in use.

Its record is mixed. Energy and storage were integrated. The Semi exists in small numbers and the bus does not. Self-driving has not been achieved. The shared fleet, which is the robotaxi concept, began in supervised form in June 2025, nine years after the plan.

Master Plan Part 3, 2023

The third plan, presented at Investor Day in March 2023, was unlike the others: a quantitative analysis of what a fully electrified global economy would require in terms of battery capacity, mining, manufacturing investment and land area, concluding that it is achievable with known technology and at an investment of approximately US$10 trillion.

It was less a corporate strategy than an argument about the feasibility of the transition, and it was criticised for containing very little about what Tesla would specifically do.

Master Plan Part IV, 2025

The fourth, published in 2025, concerns sustainable abundance and autonomy, and centres on artificial intelligence, robotaxis and Optimus rather than on vehicles.

It is the document that describes the company Tesla now says it is, and it contains substantially fewer specific commitments than the first one did.

Tesla and the transition

The broader question Tesla raises is whether a private company can accelerate an infrastructural transition that governments had failed to drive.

The evidence suggests it can, within limits.

What worked. Demonstrating demand was the decisive act. The established industry's position was that customers would not buy electric vehicles, which was true of the electric vehicles it was building and was not a fact about customers. Tesla's demonstration that a desirable electric vehicle would sell removed the argument and forced a reallocation of tens of billions of dollars of industry investment.

What also worked. Building charging infrastructure that no individual manufacturer had an incentive to build, and then opening the standard, solved a coordination problem of the kind usually requiring a regulator.

What did not work. The transition has not been driven by the product alone. Consumer incentives, emissions regulation, Chinese industrial policy and, in Norway, a tax structure that made electric vehicles cheaper than petrol ones, have each been necessary, and sales have fallen wherever incentives have been withdrawn.

What remains unresolved. Charging for people without private parking, which is a majority of urban residents in Europe; the grid capacity required; the raw material supply chains and their social consequences; and the substantial proportion of the global vehicle fleet that will be petrol for decades regardless of new sales.

Tesla's contribution has been to make the transition possible rather than to complete it, which is a considerable thing for a company to have done and is less than the company's account of itself suggests.

Final assessment

Tesla is a company whose historical importance is settled and whose future is genuinely uncertain, and the two facts are frequently confused with each other.

The historical importance is that the largest industry in the world is being rebuilt around a technology it had rejected, and that this happened because one company demonstrated that the rejection was wrong. Nothing about that judgement depends on what Tesla does next.

The uncertainty is that the company is valued as though it will do something it has not yet done, has been promising for a decade, and is being attempted simultaneously by better-funded competitors in the case of autonomy and by nobody successfully in the case of humanoid robotics.

Between those two facts sits an ordinary car company with an ageing range, compressed margins, declining volumes, a very strong charging network, a growing energy business, the best manufacturing cost position outside China, and a chief executive whose conduct has cost it a measurable and growing proportion of its customers.

Reactions to the company's products

A record of how each principal product was received when new is useful because the reactions have varied considerably and because several were wrong.

Roadster. Received as a curiosity and as an expensive toy, with reviewers impressed by the acceleration and sceptical about the price and the practicality. The scepticism was reasonable given that approximately 2,450 were sold and that the company nearly failed building them.

Model S. Received as a genuine advance, with Motor Trends unanimous Car of the Year vote in 2013 and Consumer Reportss highest score ever recorded. The criticism concerned build quality and the service network, both of which were genuine and neither of which has been fully resolved.

Model X. Received poorly relative to the others, with the doors criticised as an unnecessary complication and the reliability among the worst the company has produced.

Model 3. Received well as a product and disastrously as a production exercise, and the gap between the reviews and the delivery timetable was the defining feature of 2017 and 2018.

Model Y. Received as a competent extension of the Model 3 into the segment that dominates global sales, and became the best-selling vehicle in the world within three years, which is the commercial vindication of a product nobody described as exciting.

Cybertruck. Received as a provocation, which was the intention, and subsequently as a commercial disappointment. Its supporters regard the design as genuinely original in an industry of consensus; its critics regard it as a vehicle designed to be noticed rather than used, and the sales figures have favoured the critics.

Semi. Received well by the small number of operators using it and produced in numbers too small to assess.

Powerwall and Megapack. Received as straightforwardly good products in a market that needed them, and the Hornsdale installation in South Australia changed how grid operators think about storage.

Full Self-Driving. Received with a division that has not narrowed, between users who describe it as remarkable and improving and safety researchers who describe the name as misleading and the supervision requirement as inadequately enforced. Both descriptions have been accurate throughout.

Optimus. Received with interest and considerable scepticism, the latter reinforced by the teleoperation of the October 2024 demonstration units, which the company did not disclose at the time.

Notes on terminology

Several terms used in discussion of Tesla are used inconsistently and are worth defining.

Autonomy levels are defined by SAE International from 0 to 5. Level 2 provides steering and speed control with continuous driver supervision, which is what Autopilot and Full Self-Driving (Supervised) provide. Level 3 permits the driver to disengage under defined conditions with the system requesting takeover, which Mercedes-Benz has certified in limited circumstances and Tesla has not. Level 4 requires no driver within a defined domain, which is what Waymo operates. Level 5 requires no driver anywhere, which nobody has.

Full Self-Driving is a product name and not a level, and Tesla has added the word Supervised to it since 2024.

Range figures differ by test cycle. The American EPA cycle, the European WLTP cycle and the Chinese CLTC cycle produce progressively more optimistic results for the same vehicle, and comparisons across markets are meaningless without specifying which.

Gigafactory is Tesla's term for a plant and does not describe its function; several produce no batteries.

Battery capacity is quoted gross and usable, the latter being smaller because a buffer is reserved to protect the cells, and manufacturers differ in which they publish.

Deliveries rather than sales are Tesla's reported measure, which counts vehicles handed to customers in a quarter and which permits the timing effects that make quarterly comparisons difficult.

Battery electric vehicles are distinguished from plug-in hybrids in the comparison with BYD, which sells substantial numbers of both; BYD's total including hybrids exceeded Tesla's several years before its battery electric figure did.

Related reading

The company's history is unusually well documented for one so young, and several sources are worth identifying.

Corporate filings from the 2010 listing onward record the finances, the risks the company itself identifies, and the compensation arrangements, and the proxy statements for the 2018 and 2025 packages set out the milestones in detail.

The Delaware judgment in Tornetta v. Musk, issued in January 2024, is the most thorough public examination of the company's governance and of the relationships between Musk and his directors, and it is written in unusually plain language for a Chancery opinion.

The National Highway Traffic Safety Administration publishes its investigations, its recall notices and the standing general order data on advanced driver assistance crashes, which is the principal independent source on Autopilot's safety record.

Journalism of note includes the Wall Street Journals and Reuterss investigations into the Fremont plant's conditions and injury reporting, the New York Timess and Washington Posts reporting on Autopilot collisions, Bloombergs production tracking during the Model 3 period, and the coverage of the Swedish labour dispute in the Nordic press.

Books include Ashlee Vance's 2015 biography of Musk, Walter Isaacson's 2023 one, Tim Higgins's Power Play on the Model 3 period, and Edward Niedermeyer's Ludicrous, which is the most sustained critical treatment.

Technical sources include the company's own battery and investor day presentations, Munro and Associates' teardown analyses, which have been unusually influential in shaping the industry's understanding of Tesla's cost structure, and the published research on lithium-ion chemistry from Jeff Dahn's laboratory at Dalhousie University, with which Tesla has a long-standing research agreement.