Tesla earnings report transcript: Betting heavily on robotics, autonomous driving and energy storage, capital expenditures continue to expand
[Transcript of Tesla’s financial report: Betting heavily on robots, autonomous driving and energy storage, capital expenditures continue to expand] Tesla released its second-quarter financial results after the close of trading on Wednesday Eastern Time. Although the vehicle delivery and energy storage installation data were impressive, the decline in gross profit margin and high investment in production expansion in the quarter caused market concerns, and the stock price fell after the market closed. In a conference call after the release of the financial report, Tesla CEO Musk said that Tesla is undergoing "the fastest industrial scale expansion in the United States since World War II" and reiterated that capital expenditures this year are expected to exceed US$25 billion and will continue to grow in the next two to three years.
Tesla released its second-quarter financial results after the close on Wednesday Eastern Time. Although the vehicle delivery and energy storage installation data were impressive, the decline in gross profit margin and high investment in production expansion in the quarter caused market concerns, and the stock price fell after the market closed. In a conference call after the release of the financial report, Tesla CEO Musk said that Tesla is undergoing "the fastest industrial scale expansion in the United States since World War II" and reiterated that capital expenditures this year are expected to exceed US$25 billion and will continue to grow in the next two to three years. Tesla's management also responded to the market's core questions around its four major businesses: humanoid robots, self-driving taxis, energy storage, and self-developed chips, and announced a long-term investment strategy that will continue to increase capital expenditures. It was also disclosed that Tesla is cooperating with SpaceX in multiple dimensions while facing core risks of market concern such as supply chain, regulation, and profit pressure. Complete transcript of Tesla’s 2026 second quarter earnings conference call (AI-assisted translation, some content has been deleted) Travis Axelrod (Head of Investor Relations) Good afternoon, and welcome to Tesla's 2026 second quarter earnings call Q&A. I'm Travis Axelrod, head of investor relations, and I'm joined today by Elon Musk, Vaibhav Taneja and a number of other executives. Before officially entering the Q&A, Elon Musk will give an opening speech. Over to you, Elon. Elon Musk (CEO) Thank you all. The overall performance this quarter was very impressive, and we delivered the best second-quarter delivery results in history. Model Y is currently the best-selling model in the world. It continues to set sales records in various markets around the world, and the market popularity continues to surge. In areas where FSD fully autonomous driving has been approved for implementation, the optional installation rate is extremely high. To be honest, the core appeal of many customers when entering the store is FSD, and the vehicle is just a supporting carrier. Customers at US stores have made it clear that as long as they have FSD, it can be matched with any car model. It is obvious that FSD has become the core engine driving demand. As FSD is gradually approved in various countries around the world, we expect that demand in various places will usher in similar explosive growth, and the prospects for this area are very optimistic. The energy storage business is also growing rapidly and will become a key support for the large-scale layout of AI data centers in the future. We continue to increase investment in core business and comprehensively lay out the future track. The scale of capital expenditures this year is unprecedented, and I firmly believe that all projects invested will eventually bring extremely high returns, and may even record the best capital investment returns in Tesla's history. The execution performance of Tesla's entire team deserves high recognition. The Cybertruck has been officially put into production, and the Optimus humanoid robot production line is about to be put into production; Tesla semi-trailer trucks are now in mass production; the third-generation Megapack energy storage equipment is about to be put into production; the lithium refinery and cathode material refinery have also been put into operation; battery cell production capacity continues to increase; at the same time, we are preparing for the construction of large-scale solar cell and module production lines. The entire industrial chain covers the entire process from silicon material purification, solar cell production to terminal photovoltaic installation. In the future, transportation electrification and the artificial intelligence industry will generate massive power demand, and photovoltaic energy storage combinations will become a critical need. We are advancing the most ambitious expansion plan of advanced infrastructure and manufacturing capacity in human history, which is of great significance. The development goals of self-driving taxis (Robotaxi) are very ambitious, but we must strictly control safety risks and prevent any casualties. There are 30,000 to 40,000 fatal car accidents every year in the United States alone. Such accidents rarely make media headlines. But as long as one of our self-driving vehicles causes an injury, it will appear on the front pages of major media around the world, and regulators will immediately tighten control over our business.
We never want any injuries to occur. We are expanding our self-driving taxi business at the fastest speed controllable by humans. The bottom line is zero harm to people, and ideally even to avoid accidental injuries to pets. The only constraint on business expansion is safety standards. Every time a new operating city is added, we will update the operating dynamics in real time on the X platform. You can follow Tesla's official account to get the latest progress. We are currently operating in many cities in Florida, Texas, and the San Francisco Bay Area. The pace of subsequent expansion will continue to accelerate. The weekly mileage growth rate remains above 10%, and the compound growth rate is very impressive. Let’s talk about the Optimus humanoid robot. I have mentioned many times before that Optimus will become the largest product in human history. However, the overall research and development is extremely difficult. It is necessary to build a universal humanoid autonomous robot that can complete tasks independently through voice commands or video demonstrations without manual programming. Currently, no company in the world has achieved this. The mechanical and electrical structure design of the whole machine must take into account high flexibility, long-term stability and durability, and there are a lot of technical difficulties. If the robot has a high failure rate and the 70-kg equipment needs to be manually transported after failure, its practicality will be greatly reduced. We must eliminate the risk of robots losing control. Massive R&D resources are required to design the entire machine and ramp up mass production. I want to share my expectations with the market: Optimus’ mass production ramp-up is far more difficult than any of Tesla’s past products. All parts of the entire machine are newly developed. The higher the innovation level of the components, the longer the mass production ramp-up period. Traditional electric vehicles have at least a mature supply chain for body parts, rearview mirrors, glass and other accessories. However, Optimus does not have a ready-made supporting industry chain. We must build a supply chain from scratch, or develop and produce everything ourselves. The original Model S and Model But everyone must look at the pace of mass production rationally: the production capacity of humanoid robots will follow the classic S-shaped climbing curve of the manufacturing industry, but the curve in the early climbing stage will be very flat and the cycle will be long. The root cause is the new development of all machine parts. Most of the various robot demonstration videos on the Internet are pre-programmed or remotely controlled. No humanoid robot can truly complete common daily tasks. Optimus will be the first product to achieve this capability. It is not just a demonstration prop, but can truly solve daily work needs. Optimus benchmarking even surpasses human hand dexterity. The mechanical structure of the human hand is extremely delicate. Not only the opposable thumbs, but also the control logic of subtle movements is extremely complex. After in-depth study, you will be more and more amazed. Optimus will achieve hand operation capabilities that are equal to or even surpass those of humans, and has huge potential in the future. Next, let’s talk about the Terafab chip factory project. We will soon announce the location of the factory and disclose complete planning details. Relevant heavy-hitting information will not be released in the earnings conference call, but will be held separately at a product launch conference for detailed interpretation. This project is of great strategic significance and is a necessary prerequisite for large-scale mass production of Optimus. Otherwise, AI chip production capacity will directly limit the delivery of humanoid robots. We must simultaneously overcome the problems of the entire industry chain of storage, logic chips, and packaging to support mass production of Optimus. The relevant details will be disclosed at a special press conference. All orders for Austin R&D chip factory equipment have been placed. This R&D factory integrates the entire process of photolithography mask production, logic chips, memory chips, packaging, and chip testing. The entire production line is concentrated in the same factory, which is the only one in the world. This facility can significantly shorten the chip iteration cycle and quickly verify new chip designs. We will promote high-risk, high-return cutting-edge AI chip research and development on this production line. The core research and development directions will all serve the Optimus humanoid robot to maximize the practicality of the robot.
We simultaneously develop the Digital Optimus (Digital Optimus), which has the same logic as that of self-driving vehicles: the vehicle receives picture pixels and outputs vehicle control instructions; the same goes for humanoid robots and digital Optimus, which receive visual pictures and output operating instructions. We are fully capable of migrating self-developed AI technology for autonomous driving to general computer operation scenarios and creating digital intelligence that can autonomously control computers. This system relies on low-cost Tesla AI computing units to process real-time image tasks, using high frame rate real-time video recognition instead of simple screenshot recognition. Digital Optimus is crucial to physical humanoid robots. Robots need to autonomously operate various touch screens and computer equipment. The ability to operate general electronic equipment is the foundation. This project was jointly developed by Tesla and SpaceX. SpaceX's large model Grok serves as the top-level scheduling system and issues a task list to the digital Optimus, and the digital agent completes the entire set of operations independently. We simultaneously launched the Megapod integrated computing device, which integrates the Tesla AI computing unit and the x86 processor into a package, similar to the standardized cabinet design of the Megapack energy storage device. A single Megapod integrates a large number of AI+x86 computing power combination cabinets and can be deployed in any region around the world. Relying on distributed power resources to achieve large-scale deployment of computing power, Tesla's global supercharging stations have been equipped with 7 gigawatts of power supply capacity and are continuing to expand. A large number of Megapods can be installed at supercharging sites. Relying on distributed power to support AI computing power needs, the market space is vast. In summary, the overall roadmap for autonomous driving and humanoid robots has broad prospects and reserves a large number of blockbuster products. It is difficult to fully carry out the financial report meeting. We will hold multiple product conferences to explain in detail. I think 2026 will be one of the best years in Tesla's history, and 2027 will be even stronger. Finally, I would like to thank the entire Tesla team for their efficient execution, and thank all users and investors for accompanying us all the way. Thank you all. Vaibhav Taneja (Chief Financial Officer) Thank you. The second quarter continued the recovery trend at the end of the first quarter. Global vehicle demand fully recovered, and global deliveries hit a record high: the Americas increased by 60% quarter-on-quarter, Asia-Pacific by 27%, and Europe, Africa and the Middle East by 12%. Model Y sales have all set new records in core markets such as the Netherlands, Australia, and New Zealand. The Tesla team has created a product strategy that is cost-effective and equipped with autonomous driving capabilities, and market feedback continues to be fulfilled. At the end of the second quarter, the company's order reserve reached its peak since 2023. Therefore, major factories are fully expanding production to meet the strong demand. However, production capacity expansion is constrained by the supply chain. The core bottleneck is not only batteries, but also various electronic components. Tesla is signing long-term strategic agreements with major suppliers to open up supply chain congestion. As Elon said, FSD fully autonomous driving is the core factor driving demand for complete vehicles. Sales data shows that the core appeal of a large number of customers when purchasing Tesla is FSD. Among the vehicles delivered in North America in the second quarter, 55% of customers activated the FSD subscription service when picking up the vehicle. The total number of paying FSD users worldwide is close to 1.5 million, 55% of which are one-time buyouts, and the remaining 45% choose the subscription model. We have removed the buyout package from most markets, and FSD’s future commercial growth will mainly rely on subscription business. The gross profit margin of the automotive business after excluding carbon emission points fell from 19.2% in the first quarter to 16.3%. Supplementary explanation: There was a one-time benefit of US$230 million in the first quarter (warranty account deduction, tariff reduction), but there was no similar benefit in the second quarter; if this one-time factor is excluded, the gross profit margin of the automotive business will be basically the same. This is enough to prove that Tesla’s pricing and cost control strategy is effective in the context of rising commodity prices and rising interest rates. Interest rate discount costs will directly offset revenue. Interest rates will continue to rise in 2026, and interest discount costs will rise simultaneously, further suppressing vehicle gross profit margins. The performance of the energy storage business is naturally volatile. The pace of installation and delivery is completely determined by the customer project cycle, and the company has limited controllable space. In the second quarter, energy storage installed capacity was 13.5 GWh, an increase of 53% from the previous quarter, which was the second highest quarterly installed capacity in business history. However, the gross profit margin of energy storage fell sharply from 39.5% in the first quarter to 20.4%. The combination of multiple factors led to:
1. Provision of US$240 million in warranty reserves to deal with battery cell failures in early delivery of energy storage equipment; 2. Tariff preferences exceeding US$200 million in the first quarter will no longer be extended in the second quarter; 3. Competition in the industrial energy storage market is intensifying, and the average price of products continues to decline. In the long term, the steady-state gross profit margin of the energy storage business is expected to remain in the mid-to-low range of 20%. At present, there are sufficient reserves of energy storage orders. We are simultaneously undertaking existing stock orders, and at the same time deploying data centers and the incremental demand generated by the electrification of the whole society. The gross profit margin of after-sales and other businesses increased from 9.2% to 14.1%, setting a new record high. Core driving factors: increasing business scale, optimization of fleet operating costs, covering second-hand cars, supercharging stations, service centers, insurance and other sectors. At the same time, this sector continues to invest in infrastructure construction, paving the way for the long-term expansion of self-driving taxis. Tesla’s self-driving taxi fleet continues to expand and is currently deployed in 7 regions in the United States. The pace of fleet expansion will continue to accelerate throughout the year, and it will simultaneously expand into new cities in the United States. It has been previously disclosed that operating expenses increased quarter-on-quarter this quarter, with the increase mainly coming from R&D investment: early investment in mass production of new products such as semi-trailer trucks, Optimus, and Cybercab, new AI computing hardware costs, and litigation-related expenses. The company is currently in a large-scale investment cycle, and operating expenses will continue to rise in 2026 and beyond, with core increments coming from research and development. The company's equity holdings in SpaceX generated a market value gain of US$1 billion, and after hedging exchange losses of US$300 million and Bitcoin depreciation of US$100 million, the company still achieved positive contribution to net profit. Free cash flow turned negative this quarter. The core reason is that capital expenditures doubled from the previous quarter. The scale of capital investment will further increase in the second half of 2026, and the total capital expenditures for the whole year are expected to exceed US$25 billion. Capital expenditures will continue to rise in the next two to three years, with funds invested in the expansion of self-driving taxi fleets, Optimus humanoid robot production lines, semiconductor factories, photovoltaic manufacturing capacity, and AI computing infrastructure, in addition to the continued expansion of vehicle factories. In addition to our own cash investment, we flexibly deploy credit tools and have been approved for a total loan line of up to US$30 billion to accelerate the implementation of various projects. We believe that this strategy can help the company seize the next round of industrial change, and will continue to take into account capital utilization efficiency during the investment process. The road to large-scale industrial expansion is full of challenges, and we must boldly deploy cutting-edge tracks. The performance growth curve has non-linear characteristics and the long-term development prospects are promising. Finally, I would like to thank all Tesla employees, customers, investors and suppliers for their trust and support. grateful. The online question platform has concentrated on issues related to Optimus supply chain, Terafab chip factory, and digital Optimus. Next, we will ask Ashok Elluswamy, Vice President of AI, to unify the content related to self-driving taxis. Over to you, Ashok. Ashok Elluswamy (VP of AI) Thank you. Most of the questions on the online platform focus on the implementation and scale-up of self-driving taxis. I systematically sorted out the current progress and long-term plans. First of all, the self-driving taxi project has performed extremely well and has a zero-bleed safety record. So far, it has completed more than 380,000 miles of autonomous driving without a safety driver in six cities in two states, without any major safety accidents and only minor incidents of being scratched by other social vehicles while stationary. I would like to emphasize the safety performance of the entire system. Previously, the industry generally believed that autonomous driving must be equipped with a full set of hardware such as lidar, multi-radar, and high-precision maps. Tesla's pure vision solution used actual operating data to prove that this view is not valid: relying only on cameras, safe, comfortable, and cost-effective autonomous driving can be achieved. This zero-accident operation record fully verifies the correctness of Tesla’s full set of AI technology routes. Secondly, let’s talk about the pace of large-scale expansion. The project was launched in Austin.
We never want any injuries to occur. We are expanding our self-driving taxi business at the fastest speed controllable by humans. The bottom line is zero harm to people, and ideally even to avoid accidental injuries to pets. The only constraint on business expansion is safety standards. Every time a new operating city is added, we will update the operating dynamics in real time on the X platform. You can follow Tesla's official account to get the latest progress. We are currently operating in many cities in Florida, Texas, and the San Francisco Bay Area. The pace of subsequent expansion will continue to accelerate. The weekly mileage growth rate remains above 10%, and the compound growth rate is very impressive. Let’s talk about the Optimus humanoid robot. I have mentioned many times before that Optimus will become the largest product in human history. However, the overall research and development is extremely difficult. It is necessary to build a universal humanoid autonomous robot that can complete tasks independently through voice commands or video demonstrations without manual programming. Currently, no company in the world has achieved this. The mechanical and electrical structure design of the whole machine must take into account high flexibility, long-term stability and durability, and there are a lot of technical difficulties. If the robot has a high failure rate and the 70-kg equipment needs to be manually transported after failure, its practicality will be greatly reduced. We must eliminate the risk of robots losing control. Massive R&D resources are required to design the entire machine and ramp up mass production. I want to share my expectations with the market: Optimus’ mass production ramp-up is far more difficult than any of Tesla’s past products. All parts of the entire machine are newly developed. The higher the innovation level of the components, the longer the mass production ramp-up period. Traditional electric vehicles have at least a mature supply chain for body parts, rearview mirrors, glass and other accessories. However, Optimus does not have a ready-made supporting industry chain. We must build a supply chain from scratch, or develop and produce everything ourselves. The original Model S and Model But everyone must look at the pace of mass production rationally: the production capacity of humanoid robots will follow the classic S-shaped climbing curve of the manufacturing industry, but the curve in the early climbing stage will be very flat and the cycle will be long. The root cause is the new development of all machine parts. Most of the various robot demonstration videos on the Internet are pre-programmed or remotely controlled. No humanoid robot can truly complete common daily tasks. Optimus will be the first product to achieve this capability. It is not just a demonstration prop, but can truly solve daily work needs. Optimus benchmarking even surpasses human hand dexterity. The mechanical structure of the human hand is extremely delicate. Not only the opposable thumbs, but also the control logic of subtle movements is extremely complex. After in-depth study, you will be more and more amazed. Optimus will achieve hand operation capabilities that are equal to or even surpass those of humans, and has huge potential in the future. Next, let’s talk about the Terafab chip factory project. We will soon announce the location of the factory and disclose complete planning details. Relevant heavy-hitting information will not be released in the earnings conference call, but will be held separately at a product launch conference for detailed interpretation. This project is of great strategic significance and is a necessary prerequisite for large-scale mass production of Optimus. Otherwise, AI chip production capacity will directly limit the delivery of humanoid robots. We must simultaneously overcome the problems of the entire industry chain of storage, logic chips, and packaging to support mass production of Optimus. The relevant details will be disclosed at a special press conference. All orders for Austin R&D chip factory equipment have been placed. This R&D factory integrates the entire process of photolithography mask production, logic chips, memory chips, packaging, and chip testing. The entire production line is concentrated in the same factory, which is the only one in the world. This facility can significantly shorten the chip iteration cycle and quickly verify new chip designs. We will promote high-risk, high-return cutting-edge AI chip research and development on this production line. The core research and development directions will all serve the Optimus humanoid robot to maximize the practicality of the robot.
We simultaneously develop the Digital Optimus (Digital Optimus), which has the same logic as that of self-driving vehicles: the vehicle receives picture pixels and outputs vehicle control instructions; the same goes for humanoid robots and digital Optimus, which receive visual pictures and output operating instructions. We are fully capable of migrating self-developed AI technology for autonomous driving to general computer operation scenarios and creating digital intelligence that can autonomously control computers. This system relies on low-cost Tesla AI computing units to process real-time image tasks, using high frame rate real-time video recognition instead of simple screenshot recognition. Digital Optimus is crucial to physical humanoid robots. Robots need to autonomously operate various touch screens and computer equipment. The ability to operate general electronic equipment is the foundation. This project was jointly developed by Tesla and SpaceX. SpaceX's large model Grok serves as the top-level scheduling system and issues a task list to the digital Optimus, and the digital agent completes the entire set of operations independently. We simultaneously launched the Megapod integrated computing device, which integrates the Tesla AI computing unit and the x86 processor into a package, similar to the standardized cabinet design of the Megapack energy storage device. A single Megapod integrates a large number of AI+x86 computing power combination cabinets and can be deployed in any region around the world. Relying on distributed power resources to achieve large-scale deployment of computing power, Tesla's global supercharging stations have been equipped with 7 gigawatts of power supply capacity and are continuing to expand. A large number of Megapods can be installed at supercharging sites. Relying on distributed power to support AI computing power needs, the market space is vast. In summary, the overall roadmap for autonomous driving and humanoid robots has broad prospects and reserves a large number of blockbuster products. It is difficult to fully carry out the financial report meeting. We will hold multiple product conferences to explain in detail. I think 2026 will be one of the best years in Tesla's history, and 2027 will be even stronger. Finally, I would like to thank the entire Tesla team for their efficient execution, and thank all users and investors for accompanying us all the way. Thank you all. Vaibhav Taneja (Chief Financial Officer) Thank you. The second quarter continued the recovery trend at the end of the first quarter. Global vehicle demand fully recovered, and global deliveries hit a record high: the Americas increased by 60% quarter-on-quarter, Asia-Pacific by 27%, and Europe, Africa and the Middle East by 12%. Model Y sales have all set new records in core markets such as the Netherlands, Australia, and New Zealand. The Tesla team has created a product strategy that is cost-effective and equipped with autonomous driving capabilities, and market feedback continues to be fulfilled. At the end of the second quarter, the company's order reserve reached its peak since 2023. Therefore, major factories are fully expanding production to meet the strong demand. However, production capacity expansion is constrained by the supply chain. The core bottleneck is not only batteries, but also various electronic components. Tesla is signing long-term strategic agreements with major suppliers to open up supply chain congestion. As Elon said, FSD fully autonomous driving is the core factor driving demand for complete vehicles. Sales data shows that the core appeal of a large number of customers when purchasing Tesla is FSD. Among the vehicles delivered in North America in the second quarter, 55% of customers activated the FSD subscription service when picking up the vehicle. The total number of paying FSD users worldwide is close to 1.5 million, 55% of which are one-time buyouts, and the remaining 45% choose the subscription model. We have removed the buyout package from most markets, and FSD’s future commercial growth will mainly rely on subscription business. The gross profit margin of the automotive business after excluding carbon emission points fell from 19.2% in the first quarter to 16.3%. Supplementary explanation: There was a one-time benefit of US$230 million in the first quarter (warranty account deduction, tariff reduction), but there was no similar benefit in the second quarter; if this one-time factor is excluded, the gross profit margin of the automotive business will be basically the same. This is enough to prove that Tesla’s pricing and cost control strategy is effective in the context of rising commodity prices and rising interest rates. Interest rate discount costs will directly offset revenue. Interest rates will continue to rise in 2026, and interest discount costs will rise simultaneously, further suppressing vehicle gross profit margins. The performance of the energy storage business is naturally volatile. The pace of installation and delivery is completely determined by the customer project cycle, and the company has limited controllable space. In the second quarter, energy storage installed capacity was 13.5 GWh, an increase of 53% from the previous quarter, which was the second highest quarterly installed capacity in business history. However, the gross profit margin of energy storage fell sharply from 39.5% in the first quarter to 20.4%. The combination of multiple factors led to:
1. Provision of US$240 million in warranty reserves to deal with battery cell failures in early delivery of energy storage equipment; 2. Tariff preferences exceeding US$200 million in the first quarter will no longer be extended in the second quarter; 3. Competition in the industrial energy storage market is intensifying, and the average price of products continues to decline. In the long term, the steady-state gross profit margin of the energy storage business is expected to remain in the mid-to-low range of 20%. At present, there are sufficient reserves of energy storage orders. We are simultaneously undertaking existing stock orders, and at the same time deploying data centers and the incremental demand generated by the electrification of the whole society. The gross profit margin of after-sales and other businesses increased from 9.2% to 14.1%, setting a new record high. Core driving factors: increasing business scale, optimization of fleet operating costs, covering second-hand cars, supercharging stations, service centers, insurance and other sectors. At the same time, this sector continues to invest in infrastructure construction, paving the way for the long-term expansion of self-driving taxis. Tesla’s self-driving taxi fleet continues to expand and is currently deployed in 7 regions in the United States. The pace of fleet expansion will continue to accelerate throughout the year, and it will simultaneously expand into new cities in the United States. It has been previously disclosed that operating expenses increased quarter-on-quarter this quarter, with the increase mainly coming from R&D investment: early investment in mass production of new products such as semi-trailer trucks, Optimus, and Cybercab, new AI computing hardware costs, and litigation-related expenses. The company is currently in a large-scale investment cycle, and operating expenses will continue to rise in 2026 and beyond, with core increments coming from research and development. The company's equity holdings in SpaceX generated a market value gain of US$1 billion, and after hedging exchange losses of US$300 million and Bitcoin depreciation of US$100 million, the company still achieved positive contribution to net profit. Free cash flow turned negative this quarter. The core reason is that capital expenditures doubled from the previous quarter. The scale of capital investment will further increase in the second half of 2026, and the total capital expenditures for the whole year are expected to exceed US$25 billion. Capital expenditures will continue to rise in the next two to three years, with funds invested in the expansion of self-driving taxi fleets, Optimus humanoid robot production lines, semiconductor factories, photovoltaic manufacturing capacity, and AI computing infrastructure, in addition to the continued expansion of vehicle factories. In addition to our own cash investment, we flexibly deploy credit tools and have been approved for a total loan line of up to US$30 billion to accelerate the implementation of various projects. We believe that this strategy can help the company seize the next round of industrial change, and will continue to take into account capital utilization efficiency during the investment process. The road to large-scale industrial expansion is full of challenges, and we must boldly deploy cutting-edge tracks. The performance growth curve has non-linear characteristics and the long-term development prospects are promising. Finally, I would like to thank all Tesla employees, customers, investors and suppliers for their trust and support. grateful. The online question platform has concentrated on issues related to Optimus supply chain, Terafab chip factory, and digital Optimus. Next, we will ask Ashok Elluswamy, Vice President of AI, to unify the content related to self-driving taxis. Over to you, Ashok. Ashok Elluswamy (VP of AI) Thank you. Most of the questions on the online platform focus on the implementation and scale-up of self-driving taxis. I systematically sorted out the current progress and long-term plans. First of all, the self-driving taxi project has performed extremely well and has a zero-bleed safety record. So far, it has completed more than 380,000 miles of autonomous driving without a safety driver in six cities in two states, without any major safety accidents and only minor incidents of being scratched by other social vehicles while stationary. I would like to emphasize the safety performance of the entire system. Previously, the industry generally believed that autonomous driving must be equipped with a full set of hardware such as lidar, multi-radar, and high-precision maps. Tesla's pure vision solution used actual operating data to prove that this view is not valid: relying only on cameras, safe, comfortable, and cost-effective autonomous driving can be achieved. This zero-accident operation record fully verifies the correctness of Tesla’s full set of AI technology routes. Secondly, let’s talk about the pace of large-scale expansion. The project was launched in Austin.