Tesla Pushes Optimus Humanoid Launch to Late 2027

Tesla has moved its target date for selling Optimus robots to outside buyers.

During the Q3 2026 earnings call, Elon Musk stated that external deliveries are now expected around the end of 2027. This marks a shift from earlier, more aggressive timelines. The company is currently in early production on its high-volume line in Fremont. This facility previously built Model S and Model X vehicles. Most of the line designed for a million units a year is already installed. However, Musk said it will take at least a year to reach the target rate of 20,000 units per week. Design changes are still occurring, which slows the ramp-up process.

The main obstacles are the AI, the electromechanical design, and overall reliability. The robot must not wear out after just one month of use. An analyst asked about specific design changes, and a Tesla engineer pointed to the hand as a major challenge. Custom gears also have reliability concerns. Musk agreed that the hand is the primary issue. Fingers must be soft enough for delicate objects but tough enough to survive constant wear. Unlike human skin, robot parts cannot heal themselves. This physical limitation makes the hardware difficult to perfect.

On the software side, Tesla is still refining its AI architecture. The company does not yet have a large fleet of robots collecting data like its cars do for Full Self-Driving. To fix this, Tesla plans to build 10,000 to 20,000 robots. These units will practice tasks in a program called the Optimus Academy. Tesla also intends to put its AI5 chip into Optimus around April 2027. Musk promised a video within a month or two. It will show the robot performing tasks that are not pre-programmed. He argued that many current demos are just scripted routines that would fail in the real world.

This is not the first time Tesla has faced these issues. Reports last month noted that while production had ramped to hundreds of units a week, the robots could not generalize to new tasks. The hand and forearm contain over 100 small parts assembled by hand. Touch sensors also had reliability problems. The AI sometimes took days to learn basic tasks. Tesla’s Q3 shareholder letter now calls the current Fremont output the first production-quality bots. This implies that every unit built before this quarter was not production quality.

The timeline has slipped repeatedly. In 2024, Musk predicted internal use in 2025 and high-volume external sales in 2026. By January 2025, he said Tesla would build 10,000 robots that year. He expected several thousand to do useful work by year-end. He later admitted that none were doing so. Now, the goal is external delivery in late 2027.

Musk claimed on the call that no one is even close to Tesla on general-purpose humanoids. Competitors, however, show different progress. XPeng turned on an automated production line for its IRON robot in Guangzhou in September. More than 80% of core processes are automated. XPeng plans mass production by the end of 2026. Deliveries are set for the first half of 2027. This is a similar timeline to Tesla’s, if not slightly ahead. XPeng has not yet shipped IRON to a paying customer, though.

Boston Dynamics started producing its electric Atlas in January. All 2026 Atlas deployments are already committed, including to Google DeepMind. The company opened a training center at Hyundai’s Metaplant in Georgia in September. Atlas is learning to sequence car parts for the assembly line. Hyundai plans to deploy 25,000 robots across its plants in the coming years. Boston Dynamics also hired Milan Kovac, Tesla’s former head of Optimus, in January. The company took a pragmatic approach to hands by prioritizing sturdy grippers for industrial parts over human-like dexterity. This solves the problem Tesla says is holding it back.

Tesla faces the same core problems as its rivals. Hands must survive real work. Robots must be reliable enough to pay for themselves. AI must handle tasks it was not trained for. No one has fully cracked generalization yet. The difference is that Tesla’s competitors are not telling investors they have no competition. Based on public information, XPeng has a more automated line and a similar delivery date. Boston Dynamics is already deploying Atlas in a car factory with committed customers. Tesla may have built the most robots, but building them is not the hard part. Making them useful is. Musk’s promised video will be a key test. If it shows real, unscripted work, it will be a significant step. But Optimus timelines have slipped every year. It remains to be seen if late 2027 will hold.