Tesla delivered ~500K vehicles, up from ~367K in FY19, driven by the ramp of Model 3 and Model Y at Fremont and Gigafactory Shanghai. Automotive gross margin improved to 26% from 21% in FY19, driven by lower manufacturing costs from Shanghai localization, higher-margin Model Y introduction, and a near-doubling of high-margin regulatory credit revenue to $1.58B. The average selling price for Model 3/Y declined due to a higher mix of lower-priced Standard Range variants.

Deliveries grew 87% to ~936K as Gigafactory Shanghai ramped exports and Model 3/Y production expanded. Automotive gross margin rose to 29.3%, driven by lower per-unit costs from Shanghai localization and increased scale. This margin expansion occurred despite supply chain headwinds from semiconductors, rising raw material costs, and a 7% decline in regulatory credit revenue. The average vehicle selling price declined 6% YoY in Q3 due to mix shift toward lower-priced models, but cost reductions more than offset this.

Deliveries grew 40% to ~1.31M as new factories in Berlin and Austin began production. Despite strong volume growth, automotive gross margin declined to 28.5% from 29.3%, driven by higher per-unit costs from raw material inflation (particularly lithium), idle capacity charges from Shanghai shutdowns and new factory ramps, and a cost premium from the shift toward Model Y. Berlin and Austin added roughly $2,000-$2,500 per unit of cost headwind. Tesla raised prices during the year in response to cost inflation and strong demand, partially offsetting these pressures.

Tesla delivered a record 1.81M vehicles, up 38% YoY, with Model Y becoming the best-selling vehicle of any kind globally. However, this volume growth was achieved through aggressive price reductions that compressed automotive gross margin to 19.4% from 28.5% in FY22. Management attributed the pricing actions to macroeconomic headwinds — rising interest rates effectively increased the cost of vehicle financing, reducing affordability. Cost per vehicle declined due to lower material and freight costs, higher fixed cost absorption, and IRA manufacturing credits, but these savings were insufficient to offset the ASP reductions.

Deliveries were roughly flat at ~1.79M, declining ~1% YoY. Automotive revenue decreased 6% to $77.1B as lower ASPs from continued price reductions and incentive programs more than offset modest volume shifts. Automotive gross margin declined to 18.4%, pressured by lower ASPs and Cybertruck ramp costs, partially offset by lower material costs and a 54% increase in regulatory credit revenue to $2.8B. Management noted cost per vehicle reached its lowest level ever at below $35K in Q4, driven by raw material cost improvements. Q1 production was disrupted by the updated Model 3 ramp at Fremont and factory shutdowns at Gigafactory Berlin from Red Sea shipping diversions and an arson attack.

Deliveries declined 9% to ~1.64M as Tesla simultaneously changed over all global factories to the new Model Y in Q1, resulting in several weeks of lost production. The company also faced brand headwinds from CEO Elon Musk's involvement in government activities and the broader political environment, which management acknowledged impacted certain markets. Regulatory credit revenue declined 28% to $1.99B due to legislative changes. Automotive gross margin fell to 17.8%, primarily from the decline in high-margin regulatory credits, lower deliveries reducing fixed cost absorption, and tariff impacts, partly offset by lower material costs. Management noted the company ended FY25 with a larger backlog than in recent years.

FSD was still in early development, with Musk expressing confidence that the system would achieve safety levels exceeding human drivers within the year. Tesla began offering FSD subscription. The FSD computer (Hardware 3) was described as the most efficient inference computer in the world despite being several years old. Musk framed FSD as potentially the 'biggest increase in asset value of any asset class in history' — arguing that autonomous capability would multiply the utility of each vehicle from ~12 hours/week to ~50-60 hours/week.

FSD Beta was expanded to ~60,000 vehicles in the US by Q4. Tesla accumulated ~100M miles of FSD usage outside highways. Musk stated he was 'highly confident' that FSD would achieve safety levels better than human drivers during the year. The company also began developing Dojo, a purpose-built AI training supercomputer, described as a 'long shot' but potentially an order of magnitude more efficient than GPU clusters for neural net training.

FSD Beta was deployed to roughly 400,000 customers in North America. Tesla recognized $324M in Q4 revenue related to the general release of FSD features and held nearly $1B in deferred FSD revenue. Management began emphasizing FSD as a source of near-100% gross margin software revenue on the existing fleet. FSD still operated as 'supervised' — requiring driver attention.

Tesla launched FSD Version 12, a complete architectural rewrite using end-to-end neural networks — replacing 330,000 lines of C++ code with AI that processes photons in and outputs vehicle controls directly. Management described Tesla as having the highest AI inference efficiency of any company and began rolling V12 out to 400,000+ vehicles. Musk began discussing FSD licensing interest from other automakers and framed the FSD opportunity as potentially making Tesla 'the most valuable company in the world.'

Management announced plans to launch unsupervised FSD as a paid service in Austin by June FY25. Thousands of Teslas were already driving autonomously at the Fremont factory — navigating from end-of-line to specific parking spots without human intervention. FSD Version 13 showed continued improvement in safety metrics, and the company invested heavily in AI training infrastructure including the Cortex cluster at Gigafactory Texas. Musk projected 'millions of Teslas operating fully autonomously' in the second half of FY26.

Tesla launched its Robotaxi service in Austin in June FY25, providing paid autonomous rides with no safety driver and eventually no chase car. By Q4, the company reported over 500 robotaxi vehicles carrying paid customers between Austin and the Bay Area. Unsupervised rides accumulated over 250K miles in Austin. FSD Version 14 was deployed to customers, with management describing it as incorporating much of the Robotaxi FSD model. Active paid FSD subscriptions reached ~1.1M in Q4, up 38% YoY. Tesla announced a full transition to subscription-only FSD pricing, which management noted would temporarily impact automotive margins. Management projected expanding to dozens of US cities by year-end and began discussing allowing personal Tesla owners to add vehicles to the fleet (Airbnb-style model) in FY26.

Energy revenue grew 30% to $2.0B, driven by increased Megapack and Powerwall deployments. However, gross margin declined to 1% from 12% in FY19, primarily due to a higher mix of Solar Roof sales operating at low margins during its manufacturing ramp. Tesla deployed 3.02 GWh of energy storage products.

Energy revenue grew 40% to $2.8B. However, gross margin turned negative to -4.6%, driven by a higher proportion of Solar Roof sales with temporary manufacturing underutilization. Management acknowledged that the energy business was 'short-changed' as chips and resources were diverted to the vehicle business during the semiconductor shortage.

Energy revenue grew 40% to $3.9B with deployments reaching 6.5 GWh. Gross margin improved to 7.4% from -4.6%, driven by a higher proportion of energy storage sales (which carry higher margins than solar) and higher Megapack ASPs. Demand for Megapack exceeded supply, constrained by semiconductor challenges.

Energy storage deployments more than doubled to 14.7 GWh, and segment revenue grew 54% to $6.0B. Gross margin improved to 18.9%, driven by lower costs per MWh for Megapack, a higher mix of storage versus solar, and IRA manufacturing credits. Management noted that energy profits nearly quadrupled YoY.

Energy revenue grew 67% to $10.1B with deployments reaching 31.4 GWh. Gross margin expanded to 26.2%, driven by continued cost reductions, IRA manufacturing credits, and a higher proportion of storage sales. The segment achieved a record 30.5% gross margin in Q3 FY24. Megafactory Shanghai was completed in December and began ramping in Q1 FY25.

Energy revenue grew 27% to $12.8B with record deployments of 46.7 GWh, up 49% YoY. Gross margin expanded further to 29.8%, driven by lower raw material and manufacturing costs for Megapack, partly from the Shanghai Megafactory ramp. The segment faced headwinds from tariffs on Chinese LFP cells (contributing over $200M in quarterly tariff costs) and increasing low-cost competition. Management guided for margin compression in FY26 from competition, policy uncertainty, and tariffs, but demand remained strong with a globally diversified backlog. Tesla announced Megapack 3 and plans for a third Megafactory near Houston for FY26.

Capex was $3.2B, directed toward Model Y expansion at Fremont, Gigafactory Shanghai expansion, and construction of new factories in Berlin and Austin. Tesla raised $12.3B from public stock offerings and generated $5.9B in operating cash flow. Cash ended the year at $19.4B.

Capex more than doubled to $6.5B, primarily for construction of Gigafactory Texas and Gigafactory Berlin, plus expansions at existing factories. Operating cash flow increased to $11.5B. Tesla did not issue new equity and instead made net debt repayments. The company also invested $1.5B in Bitcoin. Cash ended the year at $17.6B.

Capex increased to $7.2B, focused on gigafactory expansions. Operating cash flow grew to $14.7B. Tesla reduced total debt to $2.1B and entered a new $5B revolving credit facility in January 2023. Cash and investments ended at $22.2B.

Capex rose to $8.9B, directed toward factory expansion and new product tooling. R&D grew 29% to $4.0B, reflecting Cybertruck pre-production costs and AI investment. Operating cash flow reached $14.9B despite margin compression, and free cash flow was $4.4B. Cash and investments ended at $29.1B.

Capex increased to $11.3B with a meaningful portion allocated to AI infrastructure, including the Cortex training cluster at Gigafactory Texas. Cumulative AI-related capex reached ~$5B. R&D grew 14% to $4.5B. The company also recognized $684M in restructuring charges, including $583M from a 10%+ global headcount reduction in Q2. Free cash flow was $3.6B and cash grew to $36.6B.

Capex moderated to $8.5B, slightly below the $9B guide. However, R&D increased 41% to $6.4B, driven by AI initiatives, Optimus, and new product programs. The company also recorded $494M in restructuring charges related to AI chip design convergence. Operating cash flow was $14.8B and free cash flow improved 74% to $6.2B. Cash and investments grew to $44.1B. Management guided FY26 capex to exceed $20B — covering Cybercab, Semi, Optimus factory, LFP cell factory, lithium refinery, new Megafactory, and AI compute — noting this does not include potential solar cell manufacturing or semiconductor fab investments.

Elon Musk stated that Optimus was 'actually the most important product development we're doing this year,' positioning it as potentially more significant than the vehicle business. The robot was presented at AI Day. Tesla leveraged its expertise in electric motors, batteries, power electronics, and AI to develop the robot, with the same inference computer used in vehicles.

Musk described the Optimus lab as looking like 'the set of Westworld' and emphasized the program's potential to 'far exceed the value of everything else that Tesla combined.' The robot was in active development with subsystem testing stands and iterating on design. Management emphasized that the key barriers were getting the robot to perform useful tasks, scaling the AI for generalized work, and manufacturing at volume.

Tesla planned to build roughly 10,000 Optimus robots in FY25, primarily for internal factory use on boring, repetitive tasks. Management described an aspiration for order-of-magnitude production ramp per year, targeting 1M+ units annually within approximately five years. At steady-state volume of 1M/year, production cost was expected to be under $20,000 per unit. External sales were tentatively projected for the second half of FY26.

Optimus robots walked around Tesla's Palo Alto offices 24/7, could guide visitors to meeting rooms, and performed basic factory tasks. An Optimus unit performed kung fu at a movie premiere autonomously. Management announced plans to replace the Model S/X production line at Fremont with a 1M unit/year Optimus 3 production line. Optimus 3 prototypes were expected in Q1 FY26, described as so lifelike that 'people could be easily confused that it is a human.' The rare earth magnet supply chain from China posed a near-term constraint on actuator production. Management emphasized that Optimus required an entirely new supply chain with nothing off-the-shelf, making the production ramp inherently unpredictable.

Regulatory credit revenue nearly doubled to $1.58B, becoming a meaningful contributor to Tesla's first full year of profitability. Management stated the company did not plan its business around credit sales.

Regulatory credit revenue declined 7% to $1.47B. Management noted these would continue to decrease in materiality going forward as the core business scaled.

Regulatory credit revenue increased 21% to $1.78B, including a one-time $288M recognition in Q1 from credits sold under a regulatory change. Excluding this, revenue was roughly flat. Tesla also recognized a $204M impairment on Bitcoin holdings.

Regulatory credit revenue was flat at $1.79B. Net income of $15.0B was significantly inflated by a one-time $5.9B non-cash tax benefit from releasing a valuation allowance on US deferred tax assets, which reduced the effective tax rate to -50%. Excluding this benefit, the underlying trajectory showed profitability declining due to vehicle price cuts.

Regulatory credit revenue increased 54% to $2.8B, which Tesla attributed to increased demand from other OEMs scaling back their BEV plans. This high-margin revenue partially offset the decline in vehicle ASPs. The effective tax rate normalized to 20%, and management noted the FY23 tax benefit would not recur.

Regulatory credit revenue declined 28% to $2.0B due to legislative changes (OBBBA) that reduced emissions penalty amounts to zero, lowering demand for credits from other OEMs. This decline was the primary driver of the 60 bps contraction in automotive gross margin. Other income swung from a $695M gain in FY24 to a $419M loss in FY25, driven by mark-to-market losses on Bitcoin holdings and unfavorable FX movements, contributing to the decline in net income. The effective tax rate increased to 27%.

Tesla introduced single-piece rear body castings for Model Y — the first time in automotive history — using the largest casting machine ever made. Battery Day outlined a plan to produce cells internally to supplement suppliers, targeting 100 GWh by FY22. Tesla emphasized it would continue purchasing from Panasonic, CATL, LG, and BYD while building its own capacity to accelerate growth beyond what suppliers could provide.

Tesla's 4680 cell pilot plant in the Bay Area began producing cells, described as large enough to rank in the top 10 battery cell factories globally despite being a pilot facility. Equipment installation for volume 4680 production began at Gigafactory Texas. Tesla continued to face semiconductor shortages, prompting the company to rewrite firmware and reduce chip counts in vehicles.

4680 cell production reached a rate of 1,000 vehicles per week by year-end. Tesla announced a 100 GWh expansion of cell production capacity at Giga Nevada. The company noted that nearly half of vehicles produced in Q1 used LFP batteries (containing no nickel or cobalt), diversifying cell chemistry.

4680 production was described as ahead of the Cybertruck ramp schedule. Tesla began ramping its cathode refinery and lithium refinery. Management noted the company was 'approaching the limits' of cost reduction within current platforms, motivating the next-generation vehicle platform with fundamentally new manufacturing technology.

Tesla confirmed it was designing its AI5 chip, described as potentially 40x better than AI4 on key metrics. The company had chip supply agreements with both TSMC (Arizona) and Samsung (Texas) for US-based production. Management began discussing the importance of domestic semiconductor manufacturing for geopolitical resilience.

Tesla's US vehicles were ~85% USMCA-compliant on a weighted average basis, providing a structural tariff advantage. Tariff impacts were approximately $400M+ per quarter across automotive and energy businesses. The lithium refinery in Texas began production, and LFP cell manufacturing lines in Nevada were being commissioned. Management announced aspirations for a 'TerraFab' integrating logic, memory, and chip packaging — citing concern that even maximum output from Samsung, TSMC, and Micron would be insufficient for Tesla's 3-4 year growth plans in autonomous vehicles and robots. The company also began producing Tesla-branded solar panels at its Buffalo factory.