Tesla’s Full Self-Driving (Supervised) v14.3.1, rolling out with software update 2026.2.9.7, introduces foundational architectural changes that significantly reduce reaction times and improve decision-making in edge cases. This release primarily focuses on backend compiler rewrites and enhanced neural network training to handle complex driving scenarios more effectively.
Core technical upgrades
The most significant change in this update is the complete rewrite of the AI compiler and runtime using MLIR (Multi-Level Intermediate Representation). This infrastructure upgrade translates neural network models into instructions for the Hardware 4 computer much more efficiently, yielding a 20% faster reaction time and quicker model iteration.
Additionally, Tesla upgraded the neural network vision encoder and the Reinforcement Learning (RL) stage, which improves the system’s 3D geometry comprehension and its ability to interpret low-visibility conditions.
Driving and parking enhancements
On the road, the update focuses on reducing common frustrations by mitigating unnecessary lane biasing and minor tailgating behaviors. Intersections have seen major improvements, with the system now handling complex stop signs and yellow lights much more naturally rather than halting prematurely.
For parking, the vehicle now displays a “P” icon on the map to indicate its intended spot before arriving and demonstrates much greater decisiveness when maneuvering into spaces.
Early footage from testers reinforces Tesla’s claims of faster AI reaction times and improved proactive safety. Testers have also praised the revamped intersection protocol as one of the most impactful user-facing improvements in this release.
Official release notes
Full Self-Driving (Supervised) v14.3.1 includes:
- Upgraded the Reinforcement Learning (RL) stage of training the FSD neural network, resulting in improvements in a wide variety of driving scenarios.
- Upgraded the neural network vision encoder, improving understanding in rare and low-visibility scenarios, strengthening 3D geometry understanding, and expanding traffic sign understanding.
- Rewrote the AI compiler and runtime from the ground up with MLIR, resulting in 20% faster reaction time and improving model iteration speed.
- Mitigated unnecessary lane biasing and minor tailgating behaviors.
- Increased decisiveness of parking spot selection and maneuvering.
- Improved parking location pin prediction, now shown on a map with a P icon.
- Enhanced response to emergency vehicles, school buses, right-of-way violators, and other rare vehicles.
- Improved handling of small animals by focusing RL training on harder examples and adding rewards for better proactive safety.
- Improved traffic light handling at complex intersections with compound lights, curved roads, and yellow light stopping — driven by training on hard RL examples sourced from the Tesla fleet.
- Improved handling for rare and unusual objects extending, hanging, or leaning into the vehicle path by sourcing infrequent events from the fleet.
- Improved handling of temporary system degradations by maintaining control and automatically recovering without driver intervention, reducing unnecessary disengagements.
Upcoming Improvements:
- Expand reasoning to all behaviors beyond destination handling.
- Add pothole avoidance.
- Improve driver monitoring system sensitivity with better eye gaze tracking, eye wear handling, and higher accuracy in variable lighting conditions.
Blind Spot Warning While Parked
If you attempt to open a door while an approaching object is detected in your blind spot (for example, a bicyclist approaching from behind) the blind spot indicator light flashes, a chime sounds, and your door will not open upon initial button press. Wait a short time and press the button a second time to override the warning.
Full Self-Driving (Supervised)
Under your supervision, Full Self-Driving (Supervised) can drive your Tesla almost anywhere. It will start from a parked position, make lane changes, select forks to follow your navigation route, navigate around other vehicles and objects, make left and right turns and park at your destination. You and anyone you authorize must use additional caution and remain attentive. It does not make your vehicle autonomous. Do not become complacent.
Full Self-Driving (Supervised) is enabled on your vehicle. To use the feature, press the Start Self-Driving button on the UI, or press the right scroll wheel button once. You can disable Full Self-Driving (Supervised) in Self-Driving Settings.
You may also like:
- Tesla reveals a massive 2026 Spring Software update for your vehicle »
- Tesla begins wide rollout of 2026.8.6.1 software update: release notes »
- Tesla’s 2026.8.6 update hints at age detection through the cabin camera »
We’ve launched an official Tesla Owners Online Forum. Join early and claim your "Founding Fuel" badge.
