Tesla's Lars Moravy noted in a Munro Live engineering discussion that the Cybercab retains a charging port so the company can use existing infrastructure. He added: “hopefully Optimus will help us with that shortly.” That is a statement of hope, not an announced deployment.
Key takeaways
- The Cybercab keeps a physical charging port in order to use existing infrastructure
- Moravy gave no date and no operational commitment for such a role
- A Tesla engineer described Optimus's battery as one module of roughly 2.5–3 kWh
- The same person noted the battery is not the robot's largest cost
- Tesla previously said Optimus production had begun on its high-volume lines in Fremont
What does autonomous driving not solve?
A Cybercab finishes a run and needs charging. It can reach the stall on its own — that part it handles. The problem starts a metre further on.
Someone has to pick up the plug and insert it into the socket. That is physical work, performed on equipment designed for a human hand, and no progress in driving algorithms removes it.
Vehicle autonomy does not automate every task needed to keep a fleet running.
That is the gap Moravy filled with Tesla Optimus — as a bridge, not a replacement for the company's longer-term plans for charging infrastructure.
The diagram isolates the one step autonomous driving does not close. The lower branch is marked unconfirmed because it reflects Moravy's remark, not a working solution.
What was said, and what was not?
The distinction matters more than the content. Moravy described retaining the charging port as a design decision driven by wanting to use what is already installed. The Optimus line came alongside it, as a comment on the transition toward more automated charging infrastructure.
No date was given. No commitment was made. Nor was there any technical description — how many robots per station, what cycle time, what error handling.
Later in the conversation vehicle servicing and delivery came up, and Munro's host sketched a broader scenario with autonomous trucks and robots delivering food. That is the host's speculation, not an announced Tesla plan.
Around 34 minutes into the recording, a Tesla battery engineer introduced as Robbie described Optimus's battery as one module, citing roughly 2.5–3 kWh and noting it was not the robot's largest cost. That conversational wording does not establish an exact capacity for a given generation, a runtime figure or a cost breakdown.
What do these names mean?
A few terms from this conversation decide how firmly the quoted numbers can be read.
- Cybercab — Tesla's planned autonomous vehicle without a steering wheel or pedals, intended for fleet service rather than retail sale.
- Charging port — the physical socket for a cable. Keeping it means the vehicle uses existing chargers instead of requiring new wireless or robotic infrastructure.
- Battery module — a set of cells forming one replaceable unit. Calling it “one module” describes pack construction, not a measured capacity.
- kWh — the kilowatt-hour, a measure of stored energy. Capacity alone does not translate into runtime, which depends on power draw for a given task.
Together these explain why 2.5–3 kWh should be treated as an approximate reference point, not a specification update.
Why not simply automate the charging itself?
A simpler route suggests itself: build a charger that moves to the car, or charge inductively. Tesla is weighing such directions, and those are what Moravy called the longer-term plans.
The trouble with an automated charger is that it requires rebuilding every station. A fleet spread across a city uses infrastructure that already exists and does not belong to the fleet operator.
A humanoid inverts that calculation. Instead of converting thousands of stalls, one machine capable of operating equipment designed for people suffices. This is the classic justification for the humanoid form itself, and a case of dexterous manipulation: grasp, sense resistance and insert the plug precisely.
There is a flip side. A robot at a charging station is one more machine to maintain, power and repair. The economics depend on how many vehicles one station can serve — a number nobody has published.
Is it live yet?
No. This is a remark in a conversation, with no date and no commitment. Moravy used conditional phrasing, and the source material explicitly separates the comment from Tesla's announced plans.
The production status is firmer. Tesla previously said Optimus manufacturing had begun on high-volume lines at its Fremont factory. As of August 2026 the third-generation, production-intent robot had not been publicly unveiled, and the company had disclosed no external customers or unit sales.
Moravy's remark therefore adds a possible use case, not evidence of a working capability. The exchange was highlighted by The Humanoid Hub.
Watch out, Amazon. Tesla’s got its eye on the delivery space. In a recent Munro Live interview, Tesla's Lars Moravy and Eric Earley shared some Optimus nuggets: - Optimus's battery is a single ~2.5 kWh module, and unlike in a car (where the pack is a quarter to a third of the Show more
Why does this matter?
The question of who plugs in the cable is more interesting than the answer. It shows that an autonomous fleet does not remove physical labour from the market but pushes it to the edges: charging, servicing, vehicle preparation. That is where physical AI has its nearest and most countable application.
What's next?
- Tesla has given no date or scope for an Optimus role in charging Cybercabs
- The third-generation, production-intent Optimus had not been publicly unveiled as of August 2026
- Optimus production has begun on high-volume lines in Fremont — the one confirmed piece of this picture
Sources
- Humanoids Daily — Tesla's Lars Moravy Floats Optimus as a Cybercab Charging Helper
- Munro Live — Tesla Cybercab Deep Dive: Inside the Engineering with Lars Moravy and Team
- The Humanoid Hub — Post on Lars Moravy's remark
