Tesla 24Q4 earning report summary
Highlights of the $TSLA Q4 Financial Meeting in 24
1. Financial data
- Total revenue: 25707 million (25182 million in 24Q3)
- Vehicle gross profit margin: 13.6% (17.1% in 24Q3)
- Operating profit margin: 6.2% (10.8% in 24Q3)
- Non-GAAP EPS: 0.73 (0.72 in 24Q3)
- Free cash flow: 2031 million (2742 million in 24Q3)
2. Production and sales
- Q4 achieved record production and delivery figures, close to 2 million/year.
- Model Y is the best-selling model in 2024. The refreshed Model Y is accepting pre-orders worldwide, and all factories will start producing the new Model Y next month.
- Several new products will be launched in 2025. We are still on track to launch a more affordable model in the first half of 2025 and will continue to expand our product lineup on this basis.
3. FSD
- For all the skeptics, I recommend that they test drive the latest FSD software. People have seen the huge improvements brought by V13 and subsequent versions, and V14 will be a more significant improvement on this basis.
- Unsupervised FSD will be launched in Austin as a paid service in June, fully autonomous driving without a human in the car. This does not mean that it will only work in Austin. Unsupervised FSD will be released in multiple regions in the United States before the end of this year.
- Thousands of cars are already automatically driven from the exit of the production line to the set parking space every day in the Fremont factory, and it will soon be expanded to the Texas factory and other factories around the world.
- To date, AI-related capital expenditures, including infrastructure, have accumulated to about $5 billion. For 2025, our capital expenditure expectations will be flat year-on-year.
- There are some major car companies that are interested in licensing Tesla FSD, but we will prioritize unsupervised self-driving across the US.
- HW3 can only run baby v13, but we are still working on it. We will upgrade the hardware for HW3 owners who have purchased FSD.
- There are some challenges in FSD training in China. China does not allow us to transfer training videos outside of China, and then the US does not allow us to train in China. We input Chinese road signs and traffic rules into the simulator to train the model.
- The entire road system is designed for passive optical neural networks, which is how the entire road system is designed and how we expect other cars to behave. Our FSD solution is a generalized solution, not a very specific dedicated solution, which is very difficult to maintain, and our competitors are doing this.
4. Optimus
- Built a Cortex training cluster next to the Texas Gigafactory and will continue to invest. The training needs of Optimus may be at least 10 times the training needs required for the car.
- By the end of this year, there will be several thousand Optimus robots in the factory doing useful tasks, boring, dangerous, repetitive tasks that humans don't want to do.
- We hope to produce an order of magnitude more Optimus per year, if not one, then half. It won't be long before we can produce 100 million of these products per year.
- Launch Optimus Version 2 sometime next year, and may provide Optimus robots to companies outside Tesla in the second half of next year.
- Prototypes are easy, mass production is hard. Tesla has the ability of real-world AI and the ability to scale manufacturing to millions of units per year, which other companies lack.
- This is a brand new supply chain, a brand new technology, and there is nothing ready to use.
- When producing 1 million units per year, the production cost of Optimus will be less than $20,000. The price of Optimus will be determined by market demand.
- Optimus has the potential to achieve more than $10 trillion in revenue. In the long run, Optimus will be the main value of Tesla.
5. Energy Business
- The Shanghai energy storage factory is about to start production, and we are building a third factory.
- Deployment volume is expected to increase by at least 50% year-on-year in 2025.