When can you buy Tesla Optimus? The milestones behind Musk’s timeline

Tesla says Optimus production has begun, but that does not establish a date when a customer can order and receive one. On the October 21 earnings call, Musk placed deliveries to outside companies around the end of 2027, with industrial users first. Home sales would follow as production scales; the call did not provide a separate household delivery date. That account comes from The Motley Fool’s transcript of the call. October 21 call transcript

What changed in October?

This is a change from July’s line installation and development builds. Tesla now describes a manufacturing start. Its October 21 deck does not quantify Optimus output or establish a customer handover. A line described as high-volume can still be making small numbers during its early ramp. Tesla Q3 update, page 6

According to the call transcript, Musk expected at least a year, possibly longer, to reach a target of 20,000 robots a week. He described remaining work on the AI, electromechanical design and operating life. The rate is a goal, not reported current output. Industrial-first delivery also answers a different question from January’s possible public-sales forecast. Neither is a completed sale. Musk’s manufacturing and customer remarks

How the earlier plans compare

At Davos on January 22, 2026, Elon Musk described a progression from simple factory tasks toward more complex industrial work later that year. He also anticipated possible sales to the public by the end of 2027, tying that outlook to reliability, safety and functionality. This was Musk’s forecast, not an announced retail delivery commitment. WEF transcript Event date

Tesla’s January shareholder update set out plans for a production-oriented Gen 3 design and a production start before the end of 2026. It also described eventual planned annual capacity of one million robots. The word “eventual” matters: that figure was not a forecast of a million robots built during 2026. Q4 2025 update, page 8

By April, Tesla described preparations for an Optimus line at Fremont and further preparations in Texas. In July, the company reported that the Model S and Model X lines had been decommissioned and Optimus lines were being installed. Initial builds were intended for training-data collection and further development in its Optimus Academy. April update, page 6 July update, pages 3 and 6

July’s development account and October’s manufacturing start belong at different points on the same timeline. Neither alone measures repeated useful work for a customer.

Keep four stages separate

These are separate checkpoints. A training robot may be necessary for reaching commercial readiness, while still being a cost of development.

The earlier dated evidence

Three questions for the next update

What was actually built? Look for completed units and a measured production rate, with a reporting period. Planned line capacity is useful context, but it answers a different question.

What work did the robots do? A useful disclosure would identify the task, how long it ran, how often people intervened and whether performance improved. A short successful sequence can demonstrate a capability without establishing dependable operation.

Who is using them? Internal development, productive internal deployment, customer trials and paid deliveries are separate stages. Movement between them would make a more informative headline than a new target that leaves the stage unspecified.

Why this distinction matters for TSLA

Optimus belongs to Tesla, so Tesla’s operating disclosures are the appropriate starting point for stock-related analysis. Even so, a robotics milestone has to be considered alongside the rest of the business and the expectations already reflected in its valuation.

For an active trader, the useful preparation is to write down which milestone an upcoming announcement could resolve. If the market is waiting for customer use and the news describes only a production line, the announcement may leave the central uncertainty intact. If it supplies verifiable evidence of repeated useful work, it answers a more substantial question.

The next useful evidence would show how many robots were accepted for a job, the work they completed and the support required. For TSLA, these operating details help distinguish a manufacturing announcement from a business producing repeat demand. No stock-price reaction or robot profit contribution is inferred here.

For a customer-side comparison, read what Figure’s robots actually did at BMW. It shows why a named task and reporting period matter more than an undifferentiated robot count.