The consumer robot graveyard: every dead companion robot, with a dated cause of death

Six robots, five makers, one recurring cause of death: the funding round that did not close. Every cloud-dependent robot on this list, from Jibo in 2018 to the $799 Moxie in November 2024, bricked when the servers went dark, and the lone survivor (Vector) lived only because a new owner kept paying the server bill.
Nobody selling you a companion robot wants you holding this pattern in your head all at once. A startup raises money on a warm demo, ships a cloud-dependent robot at a premium price, misses the next funding round, and switches off the servers. The robot on your counter, the one you named, stops working. Not because it broke. Because the company did.
Below is the consumer robot graveyard: every dead companion robot I can document, with dates and causes. The demographics have argued for this category for a decade. The demographics did not save a single robot on this list.
The graveyard, at a glance
| Robot | Maker | Launched | Died | Documented cause | Bricked when servers went dark? |
|---|---|---|---|---|---|
| Jibo | Jibo, Inc. | Nov 2017 | Assets sold Dec 2018; to NTT Mar 2020 | Failed to raise a Series B; smart speakers duplicated its function | Yes, cloud-dependent |
| Kuri | Mayfield Robotics (Bosch-backed) | Unveiled CES 2017 | Announced Jul 25, 2018; closed Oct 31, 2018 | Bosch declined further funding; general preorders never fulfilled | Early backer units only (late 2017); all deposits refunded |
| Anki (Cozmo, Vector) | Anki, Inc. | Cozmo Oct 2016; Vector Oct 2018 | Bankrupt Apr 2019, shut May 2019 | A critical financing round fell through | Partly, see the undead twist |
| Moxie | Embodied, Inc. | 2020 | Customers notified late Nov 2024 | A critical funding round fell through | Yes, cloud-dependent |
| Pepper | SoftBank Robotics / Aldebaran | Consumer sales Jun 2015 | Production paused Jun 2021 | Weak demand | Reduced without service contract |
Jibo (died 2018 to 2020)
Jibo was "the world's first family robot," built by Cynthia Breazeal's team out of MIT. Its Indiegogo campaign launched July 16, 2014 and blew past target; total funding across its rounds is commonly reported at about $73 million (the Crunchbase figure cited across the 2019 coverage). Retail launch was November 2017.
By the time it shipped, cheaper smart speakers (Alexa, Google Home, Siri) already did most of what it did. It failed to raise a Series B. Assets went to SQN Venture Partners in December 2018, then to NTT (via NTT Disruption) in March 2020. IEEE Spectrum's post-mortem put it plainly: the company consistently sold its vision for what it wanted Jibo to be rather than the reality of what Jibo was, and being cloud-dependent, units "won't be able to do much of anything" once connectivity is turned off.
The part everyone remembers: in early March 2019 (the rollout was reported March 4, 2019), owners started getting a farewell message, followed by one last dance. The full text, as The Robot Report printed it: "While it's not great news, the servers out there that let me do what I do are going to be turned off soon. I want to say I've really enjoyed our time together. Thank you very, very much for having me around. Maybe someday, when robots are way more advanced than today, and everyone has them in their homes, you can tell yours that I said hello. I wonder if they'll be able to do this." No exact server-shutdown date was ever published; the message itself only promised "soon," and with the assets sold on to NTT in March 2020, any flat "the servers died in 2019" claim is shakier than it looks.
Kuri (died 2018, barely shipped)
Kuri was a camera-equipped mobile home companion from Mayfield Robotics, a startup on the Bosch Startup Platform. It debuted at CES 2017. Bosch declined to fund it further; Mayfield announced the end and paused operations on July 25, 2018, then ceased all operations by October 31, 2018.
Kuri is the cleanest cautionary tale on the list because it barely reached buyers at all. A small first batch shipped to early backers in late 2017; the units built after that never left the building, per Mayfield's own July 2018 statement that "the Kuri robots that have been made will not ship to customers." Preorders took a $100 deposit against a $699 launch price (raised to $799 in June 2017), and every deposit came back; Mayfield's site later confirmed it in the past tense: "All pre-order deposits have been refunded to our customers."
Anki: Cozmo and Vector (died 2019, then came back)
Anki was founded in 2010 by three Carnegie Mellon roboticists (Boris Sofman, Mark Palatucci, Hanns Tappeiner). It raised about $182.5 million across its rounds. Its charming palm-sized robots, Cozmo (October 2016) and Vector (October 2018), were as close as this category got to a hit.
It went bankrupt in April 2019 and shut down the following month, May 2019 (the shutdown was announced April 29, 2019), after a critical financing round fell through, laying off nearly 200 people. The scale it reached first makes the fall sharper: about 1.5 million robots sold as of August 2018, and just under $100 million in revenue in 2017. The funding total is $182.5 million; ignore the "$200M raised" headline rounding. Cozmo launched at $180 in October 2016. Vector launched at $249.99 in October 2018.
The undead twist: Digital Dream Labs acquired the Cozmo and Vector assets in December 2019 and later shipped Cozmo 2.0 and Vector 2.0, which, per Wikipedia, "achieved less success." Vector is the rare companion robot that outlived its maker, a trick Pleo pulled first and more cleanly. It survived for exactly one reason: somebody new chose to keep paying the server bill.
Moxie (died 2024)
Moxie was a children's AI companion from Embodied, Inc., led by former iRobot CTO Paolo Pirjanian. It launched in 2020 at $1,499 with a year of service included, then a subscription that ran $59.99 per month. Pricing was restructured more than once on the way down; by October 2024 Embodied's own store listed the robot at $799.00.
Embodied notified customers in late November 2024 that it was closing (its closing FAQ was live by November 26). The cause, in the company's own words: "A critical funding round fell through." A lead investor who was prepared to close the round withdrew at the last minute. The story broke widely on December 9 and 10, 2024, which is why "early December" circulates as the death date. Nobody reliable has published how many Moxies were out there when the lights went out.
Because Moxie's conversational features ran through cloud servers, the robots were set to stop working, to be bricked, when those servers went dark. Embodied's own closing FAQ spelled it out: "Moxie relies on cloud connectivity for its core features, and it will not function once services end." Refunds were mostly off the table, per the same FAQ: "we regret that we are unable to offer refunds for Moxie purchases," with only a best-effort promise to prioritize buyers from the previous 30 days if the assets sold, and financed buyers pointed back to their payment-plan providers.
A postscript: in its final days Embodied released OpenMoxie, local server software and code meant to let the community keep Moxie running without the company's cloud, offered with "no guarantees, promises, or warranties" (AppleInsider, December 20, 2024). Vector's survival pattern, in other words, minus the buyer who pays the server bill.
Pepper (production paused 2021)
Pepper is the outlier: more commercial greeter than home companion, but the canonical "the humanoid companion market underdelivered" data point. SoftBank introduced it in Tokyo on June 5, 2014. Consumer sales began June 2015, priced under $2,000 plus service fees, and the first batch of 1,000 units sold out in 60 seconds. About 27,000 units were manufactured by June 2021.
SoftBank paused Pepper production in June 2021, citing weak demand. The robotics arm became Aldebaran under United Robotics Group, and Aldebaran went into receivership in 2025, which by Wikipedia's reckoning almost certainly closes off any future for the robot.
The through-line
Six robots, five makers, one lesson repeated: the device is only as durable as the vendor's balance sheet. Cloud dependence turns a funding miss into a mass bricking. Demographic tailwinds (aging populations, single-person households) keep the category funded but have never translated into enough willingness to pay $800 to $2,000 for a gadget that may go dark in two years. The survivors, Vector via Digital Dream Labs, plus the therapeutic outliers like Paro and Lovot, tend to be either narrowly medical or backed by a patient parent company.
And the pattern predates the cloud. Sony introduced the original Aibo on May 11, 1999 at 250,000 yen (about $2,500 at the time), and when sales opened on June 1 the first batch of 5,000 sold out online in 20 minutes. Sony announced its discontinuation anyway on January 26, 2006 as part of a profitability drive, and by July 2014 it had stopped repairing them entirely. No servers involved; corporate strategy alone can kill a robot you love. The 2018 relaunch as the new aibo makes it the other resurrection on this page, courtesy of exactly that kind of patient parent company.
If you are shopping today, that is the question to ask before the specs: what happens to this robot the day the company misses a round? For what is actually orderable now and what it really does, see what you can actually buy in 2026 and the tracker.
One email when something actually changes: a delivery verified, a price cut, a robot discontinued. No weekly filler, unsubscribe whenever.