Tutor Intelligence raises $75M for robot work priced from $14 an hour
Tutor Intelligence raised a $75 million Series B led by Andreessen Horowitz, bringing its total funding to $117 million. Its Sonny and Cassie robots are billed by the hour or by the pick, and the company's pricing page lists palletizing at $14 to $18 an hour. With no upfront purchase, that model could let sites pay for robots from operating budgets instead of capital requests.
This story was produced through MarketScale. See how Industrial IoT teams put it to work with AI Visibility (GEO).
Key takeaways
For stacking and unloading pallets, Tutor charges $14 to $18 per robot hour, with an 80-hour monthly minimum and a flat $10,000 implementation fee, according to Inc. Customers make no upfront purchase and sign no long-term contract.
Tutor's robots are working in almost four times as many facilities as a year ago, the company said. Nearly a third of organizations that received Tutor robots in the past 12 months were existing customers installing more robots.
Nearly a third of the organizations that got robots in the past year were existing customers adding to their fleets—a repeat rate worth watching alongside Tutor’s broader growth goals, which Dealroom notes are still targets rather than proof at scale.
Free workspace
Turn your Industrial IoT expertise into content.
Record interviews, organize footage, and write with AI on a free trial of the MarketScale platform for qualifying companies. No demo required, no credit card.
Fourteen dollars an hour. That's the starting rate Tutor Intelligence lists for a robot that stacks and unloads pallets.
Andreessen Horowitz led the Series B. According to the company's announcement on Business Wire, Union Square Ventures, which led the Series A, and Neo, which led the seed round, also joined. With this round, Tutor's total funding comes to $117 million. The Robot Report says Tutor Intelligence started in 2021 as a spinout from MIT CSAIL, the Computer Science and Artificial Intelligence Laboratory at the Massachusetts Institute of Technology.
For a plant or warehouse operations leader, the useful part is the price sheet. Tutor sells robot work by the hour or by the pick, the same way a site already pays for labor.
Tutor is selling into a big gap. In the blog post introducing its next-generation robots, the company says only 12% of U.S. factories and 10% of U.S. warehouses have robots, while nearly a million supply-chain jobs go unfilled. Its argument is that general-purpose robots turn automation from an industrial engineering project into a product any operator can buy. The hourly price is the plainest version of that claim.
Adding up the $14 rate
Aaron Mok, a staff reporter at Inc., found the palletizing terms on Tutor's pricing page. The robot costs $14 to $18 for each hour it runs. Customers commit to at least 80 hours a month and pay a flat $10,000 fee for implementation. They don't buy hardware or sign a long-term contract. Companies are billed per hour or per pick, which mirrors how they already pay for labor.
For operators, an 80-hour monthly minimum could sit on the same spreadsheet as a temp-labor invoice. Above that minimum, the bill grows with use.
Who signs off could change too. With no hardware purchase, a robot billed this way can sit in an operating budget instead of a capital request. For operators whose automation plans are waiting on capital approval, a robot with no long-term contract could be a lower-commitment way to test whether the work can be automated at all.
Tutor's lead investor draws the same contrast with a conventional purchase. In the firm's investment note, the Andreessen Horowitz partners write that robotics has usually been limited to large enterprises that can pay millions upfront for expensive systems and implementation. They say Tutor's hourly pricing shifts capex to opex and opens the door to small and midsize manufacturers, where much of America's manufacturing sits, as well as some of the largest logistics companies. The partners add that customers they spoke with kept bringing up the business model.
Timing is the other half of that comparison. Tutor says Cassie starts picking on the first day of most deployments, and the median time from signed paperwork to robots in production is about 40 days. The a16z partners cite implementation times as low as 8 hours, and Tutor pitches its fleets as "delivered in days not years." These figures come from the company and its investor. Still, they give a buyer something concrete to set against the approval and installation cycle of a capital project at their own site.
Part of that speed comes from the hardware. Tutor says the next-generation Cassie generates its own vacuum onboard and needs no physical install for mostly stationary jobs such as induction and palletizing. For full case picking, the company pitches it as B2B fulfillment automation for "$0 CapEx and no major facility retrofits." That could explain how a flat $10,000 fee covers implementation. If nothing gets bolted to the floor, there's less site work to price.
Before comparing $14 an hour to a wage rate, find out what puts a site at the top of the range. Tutor's terms include deployment, maintenance and upgrades in the hourly price. Ask which job conditions push the rate toward $18.
Twice the robots, almost four times the sites
All of the adoption numbers come from Tutor. Its release compares the current footprint with a year ago, and the post announcing its next-generation robots adds a faster-moving figure: Tutor says its fleet is growing 20% month over month. Taken together, twice the robots across almost four times the facilities suggests the average site runs fewer Tutor robots than it did a year ago. That fits a model built on small, low-commitment starts.
Tutor's adoption figures
The Robot Report; Business Wire
The repeat-order figure stands out. Nearly a third of the organizations that got robots in the past year were existing customers adding to their fleets. According to Inc., Josh Gruenstein, Tutor's co-founder and CEO, wrote in an essay on X that Tutor's customers range from family-owned businesses to Fortune 50 companies. He also wrote that its robots have helped make and ship "hundreds of millions of products" across North America.
Those products are everyday consumer goods: breakfast cereals, snacks, candy, cosmetics, home goods and toys. Tutor says they move through some of the country's largest manufacturing and logistics networks.
Tutor's release ties the growth to companies that have stopped funding pilots and now pay for systems that deliver measurable output in their existing operations. Andreessen Horowitz's partners call this landing and expanding as customers see value. For a consumer-goods operator sizing up the vendor, the repeat rate may be the better benchmark than the facility count. It shows whether a first robot did well enough to earn a second order.
Remote tutors and a 100-robot test floor
Tutor's case for the hourly model rests on how its robots learn. AI models learned to converse and code from the vast record of text and images on the internet, the company argues, and the physical world has no comparable record. So Tutor puts internet-connected robots on real customer jobs, where remote human operators called tutors step in with input. Every completed task produces training data, and better models let the robots take on more kinds of work.
Tutor also built a test floor for that learning. Earlier this year it opened Data Factory 1, a fleet of 100 Sonny robots collecting data to train its foundation models. The Robot Report says the company uses it to evaluate policies across 100 physical robots, and Tutor calls it the largest robot data factory in the United States. This suggests a split in roles: customer jobs supply real-world experience, and the Data Factory gives Tutor a controlled place to check model behavior on physical hardware, away from customer sites.
Inc. describes Cassie as a robotic arm on wheels that handles boxes, stacks and unloads pallets, and gathers cases to fill orders. Sonny has two arms for individual products: it picks items from shelves into order bins and delivers them to a packing area.
The two robots run different models. According to Tutor, Cassie runs its Turing foundation model series, at 200 million parameters, and uses cameras to take on new tasks without programming, operator configuration or vendor customization. Sonny will be Tutor's first deployed robot built for TI-0, a 4.5-billion-parameter vision-language-action model. Its first job is e-commerce fulfillment: moving through brownfield warehouses, picking arbitrary items from racking and delivering finished orders to packout.
The latest Cassie, announced in October, can move through a facility fully on its own. It has four wheels and enough battery to last more than a shift. On top of ISO-certified collaborative hardware and 2D LiDAR, it has two 3D LiDAR units for detecting people and navigating without collisions. Mobility affects the hourly math. A robot that can drive between workstations could log productive hours across several tasks instead of sitting idle at one cell.
In Tutor's model, a tutor is a remote human operator who gives input to a robot working a live customer job. Tutor says its robots will be built in and deployed from the U.S., and The Robot Report noted the tutors may be overseas.
That affects who needs to review a deployment. A Tutor robot on the floor is a connected device that remote people can direct. For buyers, that could mean IT and security teams reviewing network access and remote-operator arrangements alongside the hourly rate. Upgrades also run on a different schedule from a robot you buy outright. The customer is paying for a machine that keeps getting better through the jobs it works, with new capabilities pushed to it over time.
Raghu Raghuram, the Andreessen Horowitz general partner and former VMware CEO who backed the round, made a similar argument in the release. He said deployment is how robotics gets past the limits of research, and that learning continuously across a growing range of deployments is the path from specialized robots to general-purpose ones.
The investor note explains how that worked in practice. Most robotics companies, the a16z partners write, build a general model first and then look for a use case, which takes massive upfront data collection and capital. Tutor did it the other way around. It found a real job, put hardware to work with remote tutors before building a robot brain, then worked backward to a model from the deployment data. The partners say the tutors were skilled enough that customers couldn't tell their work apart from work done by employees.
For a buyer, that history explains how Tutor can deploy into production "ahead of full autonomy," in the partners' words. A person in the loop covers what the model can't do yet, while the job generates the data meant to close that gap.
Where $75 million ranks
Dealroom.co, which tracks venture deals, said the round ranks among the largest 10% of U.S. Series B deals in deep tech, per a sample of 2,808 rounds. It also pointed to recent deals that span robot "brains" and practical deployments: Mecka AI raised $60 million for data infrastructure, and warehouse-robot startup Ultra raised $62 million using a leasing model. For operators, Ultra's leasing and Tutor's hourly billing could both mean paying for robots over time instead of buying them up front.
The strongest objection is about scale. Dealroom points out that the 1,000-robot target is still a goal, not proof the economics work at scale. Tutor also hasn't published its valuation, revenue or current robot count. Supply is the other open question. "We can hardly build robots fast enough," Gruenstein said in the announcement. The hourly model only helps sites that can actually get a robot, so manufacturing output would matter as much as demand if Tutor keeps signing new facilities faster than it builds.
The round goes after that bottleneck as much as the research. According to the release, the money funds frontier foundation model training as well as the manufacturing and deployment of Sonny and Cassie, and Tutor says it is rapidly scaling robot manufacturing in the U.S. For an operator weighing a first deployment, that makes lead time a fair question to ask on a sales call: how soon can a robot arrive, compared with the roughly 40-day median Tutor cites from paperwork to production?
Sources
- Tutor Intelligence raises Series B to scale model training, robot deployments ↗
- Tutor Intelligence Raises $75 Million Series B Led by Andreessen ... ↗ · Business Wire
- This Startup Offers Robot Workers Beginning at Only $14 an Hour ... ↗ · Inc.
- Tutor Intelligence raises $75M Series B from a16z for robot workforce ↗
- Investing in Tutor Intelligence | Andreessen Horowitz ↗
- Announcing the next generation of Cassie and Sonny | Tutor Intelligence ↗
Featured companies
Your experts belong here
Every story in MarketScale Industrial IoT starts with a company putting its controls engineers, plant-floor specialists, and integration partners on the record. Buyers are already reading this topic. The only question is whose experts they find.
Plant and controls buyers research deep before contact, and your engineers get to shape that research.
About the author
The MarketScale Newsroom reports on the companies, technologies, and trends shaping 16 B2B industries. It turns primary sources and expert commentary into clear, useful coverage for the people doing the work.