Tesla is moving its Optimus V3 humanoid toward low-volume production at Fremont in summer 2026, converting the plant's Model S and X line into a robot line after the last cars came off it in early May.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source Elon Musk said production would start this summer but ramp slowly on what he called a classic S-curve.3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source The stake is whether Tesla can turn a much-hyped robot into a manufactured product on a real line, rather than a stage demo. We assess with moderate confidence that Tesla will begin building Optimus units at Fremont in the back half of 2026, and with high confidence that meaningful volume, and any external revenue, remains a 2027-and-later question.

What is actually happening

The concrete, verifiable move is a factory conversion. Tesla ran the last Model S and X vehicles off the Fremont line in early May 2026 and planned to begin Optimus production there in late July or August, a roughly four-month turnaround in which the old line is dismantled and new equipment installed.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source Musk described the pace as extremely fast and the early output as slow, with initial robots performing simple factory skills before harder tasks.3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source On the product itself, the reveal timing has slipped repeatedly, from an expected Q1 2026 unveiling to mid-year, with Tesla citing competitive copying as a reason to hold back.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source Hardware coverage from the AWE 2026 showcase in Shanghai highlighted new hands with proportions closer to a human hand.1 Teslarati 2026-06-05 Reports Optimus mass production could begin by end of 2026, Fremont targeted at up to 1 million units annually and Gigafactory Texas at 10 million, with new human-like hands shown at AWE 2026 Shanghai. Open source

The most telling fact is a subtraction. Musk declined to give a 2026 production number, a reversal from his earlier prediction of building roughly 10,000 robots in a year, and pointed to the robot's more than 10,000 unique components as the reason output is hard to forecast.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source His framing, that the line moves as fast as its slowest part, is an honest admission that a humanoid supply chain does not yet exist at scale.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source

Second order effects and the ledger

The capacity ambitions are enormous on paper: up to 1 million units a year at Fremont and 10 million at Gigafactory Texas, with Musk saying he intends a new robot design every year and Optimus 4's design complete in 2027.1 Teslarati 2026-06-05 Reports Optimus mass production could begin by end of 2026, Fremont targeted at up to 1 million units annually and Gigafactory Texas at 10 million, with new human-like hands shown at AWE 2026 Shanghai. Open source 3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source We assess with moderate confidence that these figures function as direction rather than forecast, given that Tesla will not commit to a 2026 number for a single line.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source The gap between a 1-million-unit target and an undisclosed 2026 figure is the whole story.

Who gains. Tesla gains a manufacturing narrative it controls end to end, and by deploying the first robots into its own factories it gets a captive customer and a training environment without needing an outside buyer. Musk gains time: dropping the volume target removes the specific miss he was previously judged against. Suppliers of actuators and hand components gain a marquee, if uncertain, demand signal.

Who loses. Anyone pricing Tesla on near-term Optimus revenue loses, because internal factory deployment does not book external sales, and Tesla has announced no outside paying customers.3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source That is the sharp contrast the reader should hold: rivals in the humanoid and industrial-robot space that already have paying deployments can point to revenue, while Optimus's value is still prospective. Investors expecting the stated 1-million and 10-million figures on the implied timeline are exposed to a schedule Tesla itself will not underwrite with a number.

The counter-case

The strongest case for optimism is that Tesla has done hard, fast manufacturing conversions before, and reusing a proven vehicle line rather than building greenfield is a credible way to compress time.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source If the early units genuinely do useful factory work, Tesla can iterate on real deployment data that pure-demo competitors lack, and internal use is a legitimate on-ramp, not a dodge. For the skeptical thesis to be wrong, Tesla would need the summer line to produce working robots that measurably reduce factory labor, and to convert that into an external order within a year. The counter to the counter is history: Optimus reveal dates have slipped more than once, and a 10,000-part robot with no mature supply chain is exactly the kind of program that ramps slower than promised.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source Note that several circulating V3 spec claims beyond the hands were not confirmed in the reporting reviewed here, so treat the detailed spec sheet with caution.

What to watch

  • Fremont produces units, not just a reveal. If Tesla confirms actual Optimus units coming off the converted Fremont line by the end of 2026, the ramp is real; another reveal without production would confirm the slippage pattern.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source
  • Tesla puts a number back on the table. Watch whether Musk resumes giving a specific annual production figure within 6 to 12 months. Continued refusal signals the supply chain is still the bottleneck he described.2 Electrek 2026-04-22 Reports the Gen 3 reveal slipped to mid-2026, the last Model S and X came off Fremont in early May with Optimus production beginning late July or August, and Musk declined to give a 2026 volume target citing 10,000-plus unique parts. Open source
  • First external customer. Any announced paying deployment outside Tesla's own factories within 12 months would validate the product thesis; its absence keeps Optimus an internal tool.3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source
  • Optimus 4 design lands on time. If the Optimus 4 design is completed in 2027 as stated, the annual-cadence claim holds; a slip would undercut the 10-million-unit trajectory.3 Tesery 2026-06-08 Reports Musk saying Optimus 3 production starts summer 2026 but very slow on an S-curve ramp, with Optimus 4 design due in 2027 and a new robot design targeted each year. Open source
  • Evidence of real factory work. Independent confirmation that Fremont robots perform useful tasks, not choreographed demos, is the signal that separates a manufactured product from a showcase.1 Teslarati 2026-06-05 Reports Optimus mass production could begin by end of 2026, Fremont targeted at up to 1 million units annually and Gigafactory Texas at 10 million, with new human-like hands shown at AWE 2026 Shanghai. Open source

The line conversion is the most concrete thing Tesla has done with Optimus to date. By pulling its own volume targets, Tesla has quietly reset the game from hitting a number to simply starting to build, and the next year will show whether that modesty is prudence or a preview of another slip.