Sign in Subscribe
3 min read

Ultra Raises US$50M to Rent Robots to Warehouses Instead of Selling Them

Share
Ultra Raises US$50M to Rent Robots to Warehouses Instead of Selling Them
Ultra Raises US$50M to Rent Robots to Warehouses Instead of Selling Them

Ultra, a Brooklyn startup that leases packing robots to warehouses, raised a US$50 million Series A led by Framework Ventures, with Y Combinator also investing, Fortune reported on October 9, 2026. The round brings its total funding to US$62 million, after a US$12 million seed led by Y Combinator and NextView Ventures. Ultra went through Y Combinator's Summer 2024 batch. It was founded by CEO Jon Miller Schwartz, COO Max Friefeld, CTO Oliver Ortlieb and Chief Scientist Chetan Parthiban. Schwartz, Friefeld and Ortlieb previously built Voodoo Manufacturing, a 3D printing factory startup. The company did not disclose its valuation.

Ultra does not sell its machines. Customers pay an upfront fee to have the robots installed, then a monthly fee for hardware and software support, so they avoid a large capital purchase. The robots pack, sort and kit orders at third-party logistics sites in New York, Georgia, New Jersey and Texas. Ultra says they have packed more than 500,000 orders. Schwartz told Fortune that revenue is significant and that demand has let the company raise prices, but he did not give figures. Ultra is also deepening its partnership with Physical Intelligence, a San Francisco AI lab valued at US$5.6 billion, whose models run on Ultra's robots inside live customer warehouses.

Market Context

The deal splits the robot into two businesses. Ultra builds and installs the hardware, the "body," while Physical Intelligence supplies the "brains," software that helps each robot learn a customer's setup. In return, Physical Intelligence gets real warehouse data to train its models, and other robot makers use its software too. Physical Intelligence says its π0.6 model ran at 96.4% autonomy over an eight-hour shift in a deployment video.

Ultra is betting on fixed workstation robots rather than humanoids. Its Operator robot fits in a 5-by-5-foot space, reaches shelves up to 10 feet high and, according to the company, installs in a few hours. Schwartz told Fortune that humanoids get most of the attention but tend to fall over and are hard to deploy, and he gave them about five years to scale. Ultra competes with French warehouse robotics firm Exotec, with Amazon's in-house automation and with Atoms, the logistics venture of Uber co-founder Travis Kalanick.

The Number

500,000+ Orders Ultra says its leased robots have packed for shipping across U.S. warehouses.

Regional Relevance

For the United States, Ultra tests whether robots can be sold like software. A monthly fee lowers the barrier for mid-sized logistics operators that cannot spend millions upfront on automation. It also gives AI labs like Physical Intelligence what they need most: hours of real work in real buildings. If the model holds, more of the country's warehouse automation could be rented rather than bought.

For New York, Ultra adds to a small but growing hardware scene in Brooklyn, a borough better known for consumer brands than industrial robots. Its deployments in Georgia, New Jersey and Texas follow the main e-commerce shipping corridors, where third-party logistics firms are under the most pressure to pack more orders with the same staff.

The Other Side

How big is the business, really? Ultra has not shared revenue, the number of robots deployed or how many customers it serves. The 500,000-order figure comes from the company and is not tied to a fleet size, so it is hard to judge per-robot economics.

Who owns the advantage, Ultra or Physical Intelligence? If the intelligence lives in a partner's software that other robot makers can also use, Ultra's edge rests on installation speed and service. That is real, but it is easier to copy than a proprietary AI model.

Can rental pricing survive a downturn? Monthly fees are easy for customers to cancel when shipping volumes drop. Ultra carries the cost of the hardware either way, which makes the model more exposed to a slowdown in e-commerce than a one-time sale.

Sources & Transparency

Published by:

The Editor