Skild AI company profile: valuation, funding & revenue

This Skild AI company profile covers the company's valuation, revenue, funding rounds, investors and history. Skild AI works in robotics, with a focus on foundation models for robots. It was founded in 2023 and is headquartered in Pittsburgh, Pennsylvania, United States. Its latest valuation on record is More than US$14 billion (January 2026). Every figure links to a public source, and anything the company hasn't disclosed is marked as such.

Revenue

About US$30 million — Revenue reached within a few months of 2025 (company-stated), as of 2025. Source: MarketScreener.

Funding history

Skild AI timeline

Key dates from the sources on this page, oldest first. Metric snapshots are listed after the funding events.

• 2023 — Skild AI is founded in Pittsburgh by Deepak Pathak (CEO) and Abhinav Gupta.
• 2025 — Series B: See source, valuing the company at About US$4.5B. Investors: SoftBank (lead).
• Jan 2026 — Series C: About US$1.4B, valuing the company at More than US$14B. Investors: SoftBank (lead); NVentures (Nvidia), Macquarie Capital, Bezos Expeditions, Disruptive, 1789 Capital.

Investors

Overview

Skild AI is building a general-purpose "brain" for robots: a single foundation model, called Skild Brain, that can control many different kinds of robots, from robot arms and wheeled machines to four-legged robots and humanoids, and help them perform a wide range of tasks. Rather than programming each robot for each job, Skild's goal is to give robots a shared intelligence that lets them adapt to new environments and tasks, much as large language models gave software a general ability to work with text.

Deepak Pathak, the CEO, and Abhinav Gupta founded Skild AI in Pittsburgh in 2023. Both were professors at Carnegie Mellon University, one of the world's leading robotics research centres, and have long track records in robot learning, computer vision and self-supervised learning. Gupta previously led robotics research at Meta. Skild's Pittsburgh base gives it access to CMU talent and the city's established robotics industry.

What Skild Brain does

Skild trains its model on large amounts of data from simulation, videos of humans performing tasks and real robot experience. The company says its model is "omni-bodied", meaning it can control robots with very different shapes and abilities without being retrained from scratch for each one. Demonstrations have shown robots using Skild Brain to climb stairs, recover from being pushed, handle objects they haven't seen before and keep working when parts of the robot are damaged. Early commercial uses focus on enterprise tasks such as inspection, logistics, manufacturing and security, where robots can create value before general-purpose household robots become practical.

Business model

Skild licenses its model to robot manufacturers and enterprises, integrating it with their hardware rather than building robots itself. This "software layer" strategy aims to make Skild the default intelligence for many robot makers, similar to how operating systems power many brands of computers. Skild has said it went from zero to about US$30 million in revenue within a few months in 2025, an unusually early commercial start for a robotics foundation model company.

Funding

Skild came out of stealth in 2024 with a US$300 million Series A at a US$1.5 billion valuation, backed by investors including Lightspeed Venture Partners, Coatue, SoftBank, Bezos Expeditions, Felicis, Sequoia Capital and the Amazon Industrial Innovation Fund. A SoftBank-led Series B in 2025 valued it at about US$4.5 billion. In January 2026 it raised about US$1.4 billion in a round led by SoftBank, with NVentures (Nvidia's venture arm), Macquarie Capital, Bezos Expeditions, Disruptive and 1789 Capital, at a valuation of more than US$14 billion, roughly triple its previous level. Total funding was reported at around US$2 billion. Skild said the money would scale model training and deployment.

Competition

Skild's closest competitor is Physical Intelligence, which is also building general-purpose robot foundation models. Others include Google DeepMind's Gemini Robotics, Nvidia's GR00T models for humanoids, Covariant (whose founders joined Amazon), and humanoid makers that build their own AI, such as Figure AI and Tesla. Skild's bet is that a hardware-independent model, trained across many robot types, will generalise better and reach more customers than models tied to one robot.

Skild AI at a glance

Skild AI is a Pittsburgh robotics company founded in 2023 by Carnegie Mellon professors Deepak Pathak and Abhinav Gupta. It builds Skild Brain, a foundation model for controlling many kinds of robots. It raised about US$1.4 billion in January 2026 at a valuation above US$14 billion, led by SoftBank with Nvidia participating, and said it reached about US$30 million in revenue within months in 2025.

Why robotics needs foundation models

Traditional industrial robots are extremely precise but rigid: they repeat the same motion in a controlled environment and need reprogramming when anything changes. That limits robots to a small share of physical work. Foundation models trained on huge amounts of varied data promise robots that can handle variation, such as picking up objects of different shapes, working in cluttered spaces and following natural-language instructions. The biggest obstacle is data: unlike text on the internet, there's no vast library of robot experience to learn from. Companies are tackling this with simulation, human video, teleoperation and fleets of robots collecting data in the real world.

The sim-to-real challenge

Simulation lets robots practise millions of times faster and more safely than in the real world, but simulated physics never perfectly matches reality. A model that performs well in simulation can fail when friction, lighting or object weights differ slightly in a real warehouse. Skild and its peers use techniques such as randomising simulated conditions and fine-tuning with real-world data to bridge this "sim-to-real gap". How well companies solve it largely determines how quickly robots can be deployed in new settings.

Questions for enterprises considering robots

• Which tasks are repetitive, physically demanding or hard to staff?
• Does the robot software work with the hardware you already own?
• How is safety around human workers ensured and certified?
• How is performance measured, and what happens when the robot fails?

Why Pittsburgh matters

Pittsburgh has been a centre of robotics for decades, anchored by Carnegie Mellon's Robotics Institute, founded in 1979. The city was home to early self-driving car efforts, including Uber's autonomous vehicle unit and Argo AI, and has a deep pool of engineers experienced in perception, control and field robotics. Skild's choice to stay in Pittsburgh rather than move to Silicon Valley gives it access to that talent at lower cost and close ties to university research, a model similar to how some AI labs have grown up near leading universities.

Revenue in context

About US$30 million in revenue is small compared with a valuation above US$14 billion, implying a multiple of several hundred times revenue. Investors are paying for the possibility that Skild becomes a standard software layer across millions of robots, not for current sales. That makes the valuation highly sensitive to progress in real-world deployments over the next few years.

Latest developments (October 2026)

Skild AI raised about US$1.4 billion in January 2026 at a valuation above US$14 billion, more than triple its mid-2025 level, according to Bloomberg Law. SoftBank led, with Nvidia, Macquarie, 1789 Capital and Bezos Expeditions participating. Skild is building a general "robot brain" that can run on many types of hardware. Revenue is still small: CB Insights estimates about US$30 million in 2025. See our humanoid robot companies guide.

What to watch

Watch for named enterprise customers, partnerships with robot manufacturers, revenue growth, demonstrations on humanoid robots and competition with Physical Intelligence and big tech labs. Our guide to humanoid robot companies covers the wider field.

Takeaways

Manufacturers and logistics firms may soon add flexible AI to existing robot fleets. Founders can see how academic research leadership translates into large funding rounds. Investors should weigh a striking valuation against early-stage revenue and long deployment timelines.

Sources

Every figure on this page comes from the public sources below. Where sources disagreed, we used the company's own announcement or a primary news report. Private-company data changes; check the date beside each figure.

1. MarketScreener
2. The Bridge