
Tech • AI • Robotics
Tesla is signaling that FSD v15 is the main trigger for faster robotaxi deployment from late 2026 into early 2027, while Optimus Gen 3 could enter internal training in the second half of 2026 and begin outside sales as early as the second half of 2027.
JPMorgan Chase said after an on-site visit to Tesla that the next major inflection point for robotaxi scaling depends on the release of FSD version 15 later this year. Management described the software as a step change comparable to the jump from earlier versions, with about 40% of its core modules already being tested in the small robotaxi fleet and showing favorable early results.
Analysts said robotaxi fleet growth is likely to accelerate significantly between late 2026 and early 2027. That view aligns with recent signs of broader readiness, including approval in Nevada for Tesla to operate up to 5,000 robotaxis, though scaling remains dependent on software validation rather than vehicle supply alone.
Tesla indicated it plans to limit the number of Model Y vehicles converted into robotaxis, reflecting confidence that the dedicated Cybercab can be produced at scale in the near term. The vehicle is being built around an unboxed manufacturing approach intended to cut assembly time and cost by producing major subassemblies in parallel before final integration.
Tesla estimates total ownership cost for privately used Model 3 and Model Y vehicles at roughly 60 to 70 cents per mile. Under higher robotaxi utilization, that could fall to about 50 to 60 cents per mile, far below current ride-hailing fares of roughly $2.50 to $3.00 per mile. The longer-term target for a dedicated robotaxi platform is about 30 cents per mile.
Management said the existing AI4 hardware stack is already capable of running FSD v15 and supporting unsupervised FSD. A forthcoming AI 4.5 system is expected to add roughly 10% more compute power and double memory capacity, aimed at handling larger context windows and future robotaxi model demands.
On humanoid robots, Optimus Gen 3 has reportedly completed design finalization and most of the supply chain has been secured. That matters because Tesla previously emphasized that, unlike cars, humanoid robots lack a mature component ecosystem, making supply-chain formation one of the hardest parts of the program.
The reported rollout path has three phases. First, Optimus units would enter the Optimus Academy in the second half of 2026 to learn from real-world interaction. Second, deployment would expand inside Tesla factories. Third, outside commercial sales could begin as early as the second half of 2027, an aggressive target that would mark a rapid move from prototype work to revenue generation.
Internal deployment is expected to start with repetitive and hazardous work, especially in stamping and body-in-white operations. Final assembly is seen as a longer-term target because it still requires more complex manual dexterity. Using robots first in its own plants would let Tesla collect data while avoiding third-party compliance and customer-support complexity.
The report said Optimus Gen 3 has a long-term production target of 1 million units, while Gigafactory Texas is ultimately aiming for around 10 million units of a later generation. For context, Tesla currently produces fewer than 2 million vehicles a year globally, highlighting how central robotics has become to its long-term industrial strategy.
Tesla said more customers are now visiting stores specifically to ask about FSD, indicating the software is shifting from an optional add-on to a core purchase driver. In the United States, FSD take rates on new vehicle sales were recently reported above 55%, and the company is leaning on a $99-per-month subscription model to maximize adoption and long-term retention rather than immediate pricing gains.
In Europe, Tesla is pursuing approval both at the EU level and through member states such as the Netherlands. The company said European FSD data covering about 65 million kilometers show roughly a fivefold reduction in collisions, evidence it believes can help speed regulatory clearance after repeated delays.
The latest roadmap points to software readiness as the gatekeeper for robotaxis and factory deployment as the bridge to Optimus commercialization. If those milestones hold, Tesla could move both businesses from demonstration to scaled execution within the next two years.
Explain this