Elon Musk Reveals Tesla’s Future FSD Roadmap: What to Expect in the Next Update
Following the recent sightings of driverless Robotaxis in Austin, Elon Musk took to X to share detailed insights into Tesla’s upcoming changes for Full Self-Driving (FSD) in the coming year.
Musk disclosed information about two new versions of FSD currently in development, with one set to be released soon and another promising significant advancements in FSD capabilities.
A new software version is in the works and will be integrated into the main branch shortly. We also have a more advanced model in the alpha stage with approximately 4 times the parameters, but it still requires further refinement. This version is expected to be ready for deployment in a few months.
— Elon Musk (@elonmusk) June 10, 2025
4.5x Increase in FSD Parameters
Despite the similarity of the Robotaxis in Austin to customer vehicles, they are running a newer FSD build. Musk hinted at an imminent release of this version to customers, along with details of an upcoming version featuring a 4.5 times increase in parameters and improved memory usage.
This augmentation in parameters is expected to enhance FSD’s understanding of the environment, leading to more sophisticated decision-making and path planning. The increased memory capacity will allow for a larger context size, addressing limitations previously identified.
Musk indicated that this version will be available to customers later this year, albeit in an early alpha phase. This leap in parameters surpasses the “3X increase” mentioned in the FSD V13 upcoming improvements, suggesting significant advancements in FSD capabilities.
A new software version is in the works and will be integrated into the main branch shortly. We also have a more advanced model in the alpha stage with approximately 4 times the parameters, but it still requires further refinement. This version is expected to be ready for deployment in a few months.
— Elon Musk (@elonmusk) June 10, 2025
Significance of Increased Parameters and Context Length
The surge in parameters will enable FSD to process information with greater precision, resulting in smoother decision-making and enhanced performance in challenging conditions. Extending the context length will allow the vehicle to consider a longer history while making decisions, balancing responsiveness and informed choices.
Retraining for Improved FSD
A 4.5X increase in parameters is expected to be widely available later this year. Optimizing memory bandwidth usage and maximizing performance are crucial for maintaining efficiency. The entire system will require retraining.
— Elon Musk (@elonmusk) June 10, 2025
Enhancing the AI model’s capabilities necessitates significant computational power, prompting Tesla to focus on optimizing memory usage and performance. Musk emphasized the need to retrain FSD from scratch, with plans to expand computing resources for this purpose.
Projected FSD Timeline
Musk outlined a rough timeline for the upcoming FSD enhancements, acknowledging potential delays given the nature of innovation.
Upcoming: The current FSD build deployed in Tesla’s Robotaxi fleet will be merged into the public FSD build soon. This will serve as the next major update for customers, potentially labeled as FSD V13.3.
Late 2025: The 4.5X parameter model is expected to be released to a wider audience in the coming months, likely designated as FSD V14.
This comprehensive update from Musk sheds light on Tesla’s FSD development trajectory, offering insights into the future of autonomous driving technology.
The landscape of autonomous vehicles is evolving rapidly. Autonomous cars are poised to become ubiquitous worldwide within the next 2 to 3 years.
— Elon Musk (@elonmusk) June 10, 2025
Future Licensing Plans for FSD
Musk hinted at Tesla’s long-term strategy of potentially licensing FSD technology, given its superior safety record and advanced capabilities compared to human drivers. The vision for FSD extends beyond current applications, emphasizing its potential for widespread adoption and commercialization.
Tesla’s AI/Autopilot system could outperform the best human drivers on a track today. In the near future, this competition will be obsolete.
— Elon Musk (@elonmusk) June 10, 2025
Human Reflexes vs Machine Efficiency
Elon Musk recently tweeted that human reflexes cannot match machines, highlighting the superior capabilities of artificial intelligence and automation. As Tesla’s Full Self-Driving (FSD) technology and the Robotaxi network gain widespread adoption, other companies are likely to turn to Tesla for autonomous solutions. This presents a golden opportunity for Tesla to fulfill its long-standing vision of achieving full autonomy.
With a new version of FSD being rolled out to customers and a fully revamped version on the horizon, there is much anticipation surrounding Tesla’s advancements in autonomous driving technology. As other autonomy solutions struggle to meet financial viability, Tesla stands out as a leader in the field, poised to deliver on the promise of a self-driving future.