In April 2019, one of the most funded consumer robotics companies in history locked its doors with no warning.
What happened to Anki is a story about a startup that raised $182.5 million, generated close to $100 million in annual revenue, sold over 1.5 million robots, and still ran out of money in a matter of days.
This article covers the full timeline: what Anki built, why the funding collapsed, where Cozmo and Vector ended up, and what the shutdown revealed about the economics of consumer robotics.
If you owned a Vector or Cozmo, or you just want to understand why a company this successful could fail this fast, you will find the answers here.
What Was Anki?
Anki Inc. was an American robotics and artificial intelligence startup founded in 2010 by Boris Sofman, Mark Palatucci, and Hanns Tappeiner, three PhD students from Carnegie Mellon University’s Robotics Institute.
The company was headquartered in San Francisco and built consumer robots powered by computer vision, emotional AI, and real-time object recognition. Its goal was simple: bring advanced robotics into everyday homes at a price regular people would actually pay.
Anki is completely separate from the flashcard app of the same name. The two share nothing beyond a name. Anki Inc. made physical robots. The flashcard tool is an open-source spaced repetition program with no connection to the San Francisco startup.
Marc Andreessen of Andreessen Horowitz called Anki “the best robotics startup” he had ever seen in a 2013 blog post. That was not a throwaway comment. Andreessen later sat on the company’s board (Crunchbase).
At its peak, Anki employed roughly 200 people and had offices in both the US and Europe. The robots sold in Apple Stores worldwide, which tells you a lot about where the brand sat in the market.
What Products Did Anki Make?
Anki launched 4 main products between 2013 and 2018. Each one pushed further into AI-driven interaction than the last.
| Product | Launch Year | Core Feature | Price |
|---|---|---|---|
| Anki Drive | 2013 | AI-controlled slot car racing via smartphone | ~$149 |
| Anki Overdrive | 2015 | Modular track system, upgraded AI opponents | ~$149 |
| Cozmo | 2016 | Facial recognition, emotional responses, games | $179 |
| Vector | 2018 | Always-on home robot, Alexa integration | $249 |
Anki Drive debuted at Apple’s 2013 WWDC keynote, introduced by Tim Cook himself. That kind of launch stage is rare for a startup. It set the tone for the premium positioning the company would maintain for its entire run.
Cozmo
Cozmo launched in 2016 and quickly became the company’s biggest commercial success. It used computer vision to recognize individual faces, expressed a range of animated emotions, and could play several games with its three interactive cubes.
Cozmo was the best-selling toy in its class for all of 2017, according to NPD Group data. By mid-2018, Anki had sold over 1.5 million robots total, with hundreds of thousands being Cozmo units (TechCrunch, 2018).
The robot also had a developer SDK, which attracted a secondary community of programmers building custom behaviors. That SDK community later became important after the shutdown.
Vector
Vector launched in October 2018 at $249 and represented everything Anki had learned from Cozmo. It was always-on, cloud-connected, and designed to live on a desk rather than be put away after play sessions.
Key capabilities:
- Autonomous navigation using ultrasonic sensors and a camera
- Face recognition for individual household members
- Amazon Alexa integration for voice queries
- Emotional AI that changed behavior based on interactions
- Self-charging via a dedicated dock
Vector received a CES 2019 Innovation Award in the Robotics and Drones category. The timing was brutal. The award came months before the company closed.
How Much Funding Did Anki Raise?
Anki raised $182.5 million across multiple funding rounds from investors including Andreessen Horowitz, Index Ventures, Two Sigma Ventures, JP Morgan, C4 Ventures, and Silicon Valley Ventures (Crunchbase).
The company reported revenue approaching $100 million in 2017 and expected to exceed that figure in 2018 (Crunchbase News, 2019). For a consumer robotics company, those numbers were remarkable. No comparable company had come close.
Notable funding events:
- A $17 million round closed as recently as September 2018
- A $28 million round opened in August 2018 but never closed (Voicebot.ai, 2019)
- In 2018, investor board changes signaled internal financial strain. Marc Andreessen and Danny Rimer left director roles, replaced by representatives from JP Morgan and Two Sigma (The Robot Report)
That unclosed $28 million round is where things started breaking down. The company needed it to bridge to its next product cycle. It never arrived.
When Did Anki Shut Down?
On April 29, 2019, Boris Sofman called all staff to an all-hands meeting and announced the company was closing immediately. Employment ended Wednesday of that same week.
There was no public warning. No restructuring announcement. No gradual wind-down. Employees received one week of severance, confirmed by Recode’s reporting at the time.
The closure followed a very specific sequence:
- August 2018: $28 million funding round opened
- Late 2018 to early 2019: Negotiations with strategic investors ongoing
- April 2019: Final deal collapsed at a late stage
- April 29, 2019: All-hands meeting, layoffs announced
- May 2019: Operations fully ceased
- December 2019: Assets acquired by Digital Dream Labs
The shutdown came days after Vector received that CES Innovation Award. The company had $100 million in annual revenue and had sold millions of units. That made the collapse feel sudden to almost everyone outside the company.
Why Did Anki Shut Down?
Anki shut down because a late-stage funding deal with a strategic investor collapsed, leaving the company without the cash to continue operations. That is the direct cause. The structural causes go deeper.
An Anki spokesperson confirmed: “A significant financial deal at a late stage fell through with a strategic investor and we were not able to reach an agreement” (Recode, 2019). Acquisition talks with Microsoft, Amazon, and Comcast also failed to produce a deal (Failory).
The Failed Funding Round
The $28 million round was not a luxury. It was the bridge between current operations and the next phase of product development. Without it, the company could not sustain its engineering headcount or manufacturing costs.
Only 24% of hardware startups ever raise second-round funding, and 97% eventually die or become zombie companies (MacroFab, 2024). Anki beat those odds for years. But the consumer robotics hardware model made each funding round existential.
JP Morgan and other potential investors reportedly declined to commit, citing concerns about the long-term sustainability of the consumer robotics market as a whole.
Hardware Economics in Consumer Robotics
Hardware margins are brutal at sub-$300 price points. Cozmo sold for $179. Vector for $249. The bill of materials for robots with computer vision, emotional AI, and real-time navigation at that price range left almost nothing per unit.
The core economic problem:
- One-time hardware sales with no recurring revenue stream
- High manufacturing cost for sensors, processors, and actuators
- Continuous engineering costs for software updates and new features
- Customer support for a physical product sold to mass consumers
42% of hardware startups fail due to protracted development timelines, and 34% due to poor product-market fit (MacroFab, 2024). Anki had product-market fit. The problem was the economics of delivering on it at scale without a software revenue layer.
Jibo and Kuri (Mayfield Robotics) hit the same wall within the same 12-month window. Jibo shut down in 2018. Kuri followed in August 2018. Anki was simply the most successful version of a model that did not work financially.
What Happened to Anki’s Employees After the Shutdown?
Waymo hired Boris Sofman along with 12 other former Anki engineers in June 2019. Sofman became Director of Engineering and Head of Trucking at Waymo, leading the autonomous trucking program (The Robot Report, 2019).
The group included 5 PhD engineers from Carnegie Mellon, Stanford, MIT, and Georgia Tech, described by Axios as “an all-star engineering team.” Sofman had written his CMU thesis on autonomous driving, so the move was a return to his original research focus.
Where the founders went:
- Boris Sofman: Joined Waymo as Director of Engineering, Head of Trucking
- Mark Palatucci: Joined Waymo alongside Sofman
- Hanns Tappeiner: Hired by Apple as Director of Product Development, Special Projects Group (August 2019)
Apple held an informal career fair at Anki’s office within days of the closure announcement. Google, Microsoft, and Sonos were also present, according to a former Anki employee (Robots.Jobs, 2019). The talent dispersal was fast. Companies knew exactly what they were getting.
What Happened to Cozmo and Vector After Anki Closed?
In December 2019, Pittsburgh-based edtech startup Digital Dream Labs acquired all of Anki’s assets, including patents, trademarks, brand, customer data, and product blueprints. The acquisition was handled through Hilco Streambank and included 45 issued utility patents (35 US patents).
The acquisition price was not disclosed but was described openly by DDL founder Jacob Hanchar as posing “no risk” to Digital Dream Labs, implying a very low figure relative to the assets involved.
Digital Dream Labs Acquisition
DDL’s stated plan:
- Restore and maintain Vector’s cloud servers
- Relaunch Overdrive, Cozmo, and Vector for sale
- Introduce a tiered subscription model for Vector features
- Develop Vector 2.0 with hardware and software upgrades
DDL reached Anki’s 6.5 million existing customers directly, introducing themselves as the new stewards of the products. Hanchar described the business model shift clearly: recurring subscription revenue, not one-time hardware sales.
That was the exact model Anki had never built. DDL recognized it immediately.
Current Status of Vector and Cozmo Support
Vector remains operational under DDL ownership in 2025, though it now requires a paid subscription for full cloud functionality. The last major software update in late 2024 added multilingual support and ChatGPT integration (Loona Blog, 2025).
For users who prefer not to depend on DDL’s servers, the community-built Wire Pod project offers a local server replacement that runs Vector entirely offline. The OSKR (Open Source Kit for Robots) program, released by DDL, also allows developers to modify Vector’s firmware directly.
Cozmo development did not resume at the same pace. DDL’s crowdfunding commitments for Cozmo 2.0 faced repeated delays, drawing consistent criticism from backers. Some Cozmo units remain functional through the original app, while the developer community has maintained partial open-source alternatives for the SDK.
What Happened to Anki’s Intellectual Property?
Silicon Valley Bank held a security interest in Anki’s copyrights, patents, and trademarks from March 30, 2018, used as collateral against a loan. When Anki ceased operations, SVB had the right to seize and liquidate that IP (The Robot Report, 2019).
The physical assets inside Anki’s 40,000-square-foot San Francisco office went to auction in June 2019. The patent portfolio followed separately in November 2019, handled by Hilco Streambank.
What the IP portfolio included:
- 45 issued utility patents, including 35 US patents
- Patents, trademarks, and IP for Overdrive, Cozmo, and Vector product lines
- Territories covering US, EU, China, Germany, Canada, Japan, and South Korea
- Pending patents covering next-generation robot designs
- Customer data, social media accounts, and the anki.com domain
IP law firm Fisher & Richardson filed a lien against Anki on June 3, 2019, for unpaid patent prosecution services totaling $84,960.02 across 79 invoices dating back to August 2017 (The Robot Report, 2019). That detail alone reveals how stretched the company’s finances were, well before the public shutdown announcement.
Digital Dream Labs won the auction as the highest bidder in December 2019. DDL founder Jacob Hanchar wrote publicly that “we did not pay much” and that the biggest ongoing cost was inheriting Anki’s third-party cloud service licensing agreements (The Robot Report, 2019).
Why Did Consumer Robotics Fail to Scale in Anki’s Era?
Anki, Jibo, and Kuri collectively burned through roughly $300 million in combined funding and all shut down within 12 months of each other (RoboLabs, 2025). That is not a coincidence. It reflects a structural problem with the consumer robotics model as it existed between 2015 and 2019.
| Company | Shutdown | Fatal flaw |
|---|---|---|
| Jibo | Late 2018 | $900 price, no utility beyond novelty |
| Mayfield Robotics (Kuri) | August 2018 | Parent company Bosch needed meaningful revenue |
| Anki | April 2019 | Zero recurring revenue, high engineering burn rate |
Why One-Time Hardware Sales Could Not Sustain the Model
Fatal flaw: Zero recurring revenue. Once a customer bought the robot, the revenue stream ended. But Anki kept paying for cloud infrastructure and 200-plus engineers (RoboLabs, 2025).
Anki achieved the revenue of a hardware hit but carried the burn rate of a software company building AI systems. Those two things do not balance at a $249 price point.
For Jibo to justify its $70 million in VC funding, it needed to sell at least 100,000 units per year at its $900 price to generate acceptable returns, according to IEEE Spectrum’s analysis. The market simply did not exist at that volume.
Why iRobot Survived While Others Did Not
iRobot’s Roomba worked because it solved a recurring domestic problem, floor cleaning, rather than providing entertainment or companionship. Customers did not stop needing clean floors after the novelty wore off.
Companion robots faced the opposite dynamic. Novelty drove initial purchases. Sustained daily use required genuine utility that none of the 2015-2019 robots could fully deliver at their price points.
The consumer robotics market was valued at USD 13.4 billion in 2025, projected to reach USD 116 billion by 2035 (Future Market Insights, 2025). That growth is happening. But it is being driven by robot vacuums, lawn mowers, and functional home automation, not companion robots.
What the Market Needed That Anki Did Not Build
Subscription revenue. Specifically, a software layer that customers paid for monthly, giving the company ongoing income to fund engineering without depending on new hardware sales cycles.
What was missing:
- No app store or paid skills ecosystem
- Free software updates reduced perceived need to buy new hardware
- No enterprise or educational tier to supplement consumer sales
Digital Dream Labs understood this immediately. Their first structural change after acquiring Anki’s assets was building exactly the tiered subscription model Anki never had.
What Is the Current Status of Vector in 2025?
Vector remains operational under Digital Dream Labs in 2025, but its status is complicated. DDL released new stock as recently as January 2025, and the late 2024 software update added multilingual support and ChatGPT integration (Keyiro, 2025).
The robot is not discontinued. But it requires a subscription to function at full capacity.
The Subscription Model and What It Locks
Without an active subscription, Vector drops to a limited “Lite Mode.” Voice commands, Alexa integration, weather queries, and AI responses all go offline. The robot becomes, as one user described it, “a rolling clock with eyes” (Keyiro, 2025).
Subscription tiers (DDL):
- Basic plan: limited cloud features
- Standard plan: full voice commands and Alexa
- Premium plan: expanded AI capabilities and priority support
Vector 2.0, DDL’s hardware refresh, sells for $349.99 and includes a bundled free trial subscription.
Community Alternatives for Existing Owners
The Wire Pod project gives Vector owners a fully offline local server replacement. It runs on a Raspberry Pi or similar hardware and removes the dependency on DDL’s cloud entirely.
DDL’s OSKR (Open Source Kit for Robots) program lets developers modify Vector’s firmware directly, enabling custom behaviors without a subscription. Both tools emerged from the developer community that originally built around Anki’s SDK.
In August 2023, DDL’s servers experienced significant downtime that affected all Vector owners simultaneously. The outage accelerated adoption of Wire Pod and OSKR among users who wanted server independence (VectorRobot.info, 2023).
For anyone buying used, community resources at failed startups research consistently flag the Anki case as a cautionary example of what happens when a hardware product depends entirely on a third-party cloud that could disappear. That lesson applies directly to every secondhand Vector purchase made today.
FAQ on What Happened To Anki
Why did Anki shut down?
A late-stage funding deal with a strategic investor collapsed in April 2019. Without that capital, Anki could not cover operating costs for its 200-person team or continue hardware manufacturing. The company closed within days of the deal falling through.
When did Anki go out of business?
Anki announced its closure on April 29, 2019. CEO Boris Sofman told employees at an all-hands meeting that same day. Operations fully ceased by May 2019, less than a week after the announcement.
How much money did Anki raise before closing?
Anki raised $182.5 million from investors including Andreessen Horowitz, Index Ventures, Two Sigma Ventures, and JP Morgan. Despite that funding and nearly $100 million in annual revenue, the company ran out of cash when its final round failed to close.
What happened to Cozmo and Vector after Anki closed?
Digital Dream Labs acquired Anki’s assets in December 2019. Both robots were relaunched under DDL ownership. Vector now runs on a cloud subscription model. Cozmo development resumed more slowly, with community open-source alternatives filling some of the gaps.
Who bought Anki’s assets?
Pittsburgh-based edtech startup Digital Dream Labs purchased Anki’s full IP portfolio through a Hilco Streambank auction. The acquisition included 45 utility patents, trademarks, customer data, product blueprints for Cozmo, Vector, and Overdrive, and the anki.com domain.
Where did the Anki founders go after the shutdown?
Boris Sofman and Mark Palatucci joined Waymo, with Sofman leading the autonomous trucking program. Hanns Tappeiner became Director of Product Development at Apple’s Special Projects Group. All three moved into advanced robotics and AI roles within months of the closure.
Can you still use a Vector robot today?
Yes. Vector remains functional under Digital Dream Labs in 2025. Full features require a paid subscription. Users who prefer no subscription can run the community-built Wire Pod local server instead, which operates Vector entirely offline without DDL cloud dependency.
Did Anki go bankrupt?
Anki filed for bankruptcy after ceasing operations in May 2019. Silicon Valley Bank held a security interest in Anki’s IP as loan collateral since March 2018. The bankruptcy process led to the IP auction that Digital Dream Labs eventually won in December 2019.
How many robots did Anki sell before shutting down?
Anki sold over 1.5 million robots total by mid-2018, including hundreds of thousands of Cozmo units. Cozmo was the best-selling toy in its class for all of 2017. Vector launched in October 2018 and added to those numbers before the April 2019 shutdown.
Was Anki the only consumer robotics company to fail around that time?
No. Jibo shut down in late 2018, and Mayfield Robotics closed Kuri in August 2018. All three companies collapsed within 12 months of each other, each facing the same core problem: high hardware costs with no recurring software revenue to sustain operations long-term.
Conclusion
This conclusion is for an article presenting the full story of the Anki robotics shutdown, from a $182.5 million funded startup to a bankruptcy auction in under a decade.
Boris Sofman, Mark Palatucci, and Hanns Tappeiner built something genuinely impressive. Cozmo’s facial recognition, Vector’s autonomous navigation, and the emotional AI behind both products were ahead of their time.
But impressive technology without a recurring revenue model is just expensive hardware waiting for a funding round that might not come.
Digital Dream Labs kept Vector alive. The Wire Pod community kept it independent. And the consumer robotics market kept growing, just not in the direction Anki had bet on.
The lesson here is less about robots and more about burn rates, hardware margins, and what happens when a late-stage deal falls apart on a Tuesday.
Many of his resources are available on various design marketplaces and for free on Codepen.
Over the years, he's worked with a range of clients and contributed to design publications like Design Your Way, Designmodo, WebDesignerDepot, WPDean, Speckyboy, and Slider Revolution among others.
- How to Set Up Subscriptions on Google Play (Developer Guide) - July 12, 2026
- Why Work With a CMS Development Company for a Secure and Scalable Website? - July 12, 2026
- ADB Commands Cheat Sheet - July 11, 2026
