Tesla delays Optimus deliveries to late 2027

Tesla will begin delivering its Optimus humanoid robots to external customers by the end of 2027, according to Elon Musk. This delay comes as the company faces the same challenges as other humanoid robot makers: improving the dexterity of the robot’s hands, enhancing its durability, and refining its AI capabilities. Musk explained that these issues still need to be resolved before the robots are ready for general use.

Tesla has started early production of Optimus on its high-volume line in Fremont, where Model S and Model X used to be built. However, most of the line is still being installed, and it will take at least a year, possibly longer, to reach a production rate of about 20,000 units a week. This is because the design of the robot is still being finalized.

Musk listed the main issues that need to be solved before Optimus can be delivered: the AI, the electromechanical design, and the reliability of the robot. He expects these issues to be resolved around the end of next year, which is when Tesla will begin delivering Optimus to companies outside of Tesla. The company plans to start with industrial jobs that are dangerous, repetitive, or need to run around the clock.

One of the hardest hardware areas that still needs work is the hand. The hand needs to be soft enough to handle delicate objects while also being able to withstand constant wear and tear. Robots cannot heal like human skin can, so the durability of the hand is a major challenge for Tesla.

On the AI side, Musk said that Tesla is still working through problems with the AI architecture. Unlike the company’s electric cars, which are able to collect data for FSD, Tesla is still building a fleet of robots that will collect data for Optimus. To help with this, Tesla plans to build 10,000 to 20,000 robots that will practice tasks in what is called an Optimus Academy.

Musk also promised an Optimus video within a month or two showing the robot doing tasks that are not pre-programmed. He argues that many humanoid demos look impressive but only run scripted routines that wouldn’t hold up in the real world.

The timeline for Optimus has been slipping every year since 2024. In 2024, Musk said Tesla would build Optimus for internal use in 2025 and hoped to deliver it in high volume to other companies in 2026. In January 2025, he said Tesla would build about 10,000 robots that year, with “several thousand” doing useful work by year-end. He later admitted none of them were.

The same problems that were reported last month are still being faced by Tesla. The hand and forearm had over 100 small parts assembled by hand, and the touch sensors had reliability issues. The AI could take days to learn basic tasks, which slowed down the development process.

The robots that are coming off the Fremont line are now called “the first production-quality bots that we have manufactured.” So every Optimus before this quarter was not production quality. The delay in delivery is due to these ongoing challenges.

Musk said on the call that, to the best of his knowledge, “there is no one that is even close to Tesla” on a general-purpose humanoid robot. However, this is hard to square with what competitors have shown.

XPeng turned on an automated IRON production line in Guangzhou in September, with more than 80% of core processes automated. IRON has 21 degrees of freedom in each hand and runs XPeng’s physical AI model on three in-house Turing chips. XPeng plans mass production by the end of 2026 and deliveries in the first half of 2027 starting with stores and malls. That’s a similar timeline to Tesla’s, if not slightly ahead.

XPeng is dealing with the same reality as Tesla. It hasn’t shipped IRON to a paying customer, and its timelines have slipped like everyone else’s.

Boston Dynamics started producing its electric Atlas in January and says all of its 2026 Atlas deployments are already committed, including to Google DeepMind, whose AI models Boston Dynamics is integrating. In September, it opened a robot training center at Hyundai’s Metaplant in Georgia, where Atlas is learning to sequence car parts for the assembly line. Hyundai plans to deploy 25,000 robots across its plants over the next few years.

Boston Dynamics also seems to have taken a more pragmatic approach to hands, prioritizing sturdy grippers for industrial parts over human-like dexterity. That’s less impressive in a demo, but it’s a way around the problem Tesla says is holding it back.

Electrek’s Take: Anyone shocked? My prediction is that Optimus will forever be delayed. It's like FSD in Tesla cars — never quite good enough to really take off and always one or two releases away from fixing everything. At least a Tesla car with a broken FSD is still a good car. Optimus with a bad brain is an interesting-looking boat anchor.

Tesla is facing the exact same problems as everyone else: hands that survive real work, robots reliable enough to pay for themselves, and AI that can handle tasks it wasn’t trained for. Nobody has cracked generalization yet, Tesla included. The difference is that Tesla’s competitors are not telling investors they have no competition.

Based on what’s public, XPeng has a more automated production line and a similar delivery date, and Boston Dynamics is already deploying Atlas in a car factory with committed customers. Tesla may well have the most robots built, but building robots isn’t the hard part. Making them useful is.

Tesla has promised a video in a month or two. If it shows Optimus doing real, unscripted work, great. But we’ve heard about Optimus timelines before, and they’ve slipped every year. Why should late 2027 be different?