Robotics AI Attracts Record Investment as Valuations Soar
Skild AI has closed a massive $1.4 billion Series C✦ funding round, valuing the three-year-old robotics software company at over $14 billion. The round was led by SoftBank and Nvidia, with participation from Samsung, Mirae Asset, and LG.
The Pittsburgh-based startup has seen its valuation more than triple from $4.5 billion just seven months ago. Unlike traditional robotics firms that build physical machines, Skild AI focuses on creating software "brains" that can control any robot, regardless of manufacturer or design.
Asian Investors Lead the Charge
The funding round highlights Asia's growing appetite for AI robotics investments. SoftBank's Vision Fund spearheaded the deal, whilst South Korean giants Samsung, Mirae Asset, and LG all participated as strategic investors.
"Skild AI is building foundational technology for Physical AI across robots, tasks, and environments. We're proud to partner with Deepak, Abhinav, and the Skild AI team to bring that shared vision into real-world applications worldwide," said Dennis Chang, Managing Partner at SoftBank Investment Advisers.
This represents SoftBank's latest major AI bet following its USD 30 billion partnership with OpenAI for Asian data centres. The Japanese conglomerate has been particularly aggressive in backing physical AI companies, recently acquiring ABB Robotics for $5.4 billion.
By The Numbers
- Skild AI raised $1.4 billion in Series C funding, valuing the company at over $14 billion
- The startup has amassed a robotics dataset 1,000 times larger than competitors through simulations
- Company revenue grew from zero to $30 million in just a few months in 2025
- Total capital raised by Skild AI now exceeds $2 billion across all funding rounds
- Valuation increased more than threefold from $4.5 billion in seven months
Robot-Agnostic AI Models Drive Industry Shift
Skild AI's approach represents a fundamental shift in robotics development. Rather than building robots, the company creates universal software that can operate across different hardware platforms. This robot-agnostic model allows manufacturers to deploy the same AI across various machine types.
The company's "Skild Brain" platform can control robots performing complex tasks like clearing dishes, navigating stairs, and handling delicate objects. The software learns from millions of hours of internet video and simulation data.
"The surge in investment in foundational robotics models signals a clear recognition of their potential to revolutionise various industries, from manufacturing to logistics and even domestic assistance," noted industry analyst Sarah Chen from McKinsey Global Institute.
Competitive Landscape Heats Up
Skild AI joins a crowded field of well-funded robotics AI companies attracting massive investments:
- Physical Intelligence raised $600 million at a $5.6 billion valuation, led by CapitalG
- Figure secured over $1 billion in funding, pushing its valuation to $39 billion
- 1X reportedly sought up to $1 billion at a $10 billion valuation
- Boston Dynamics continues advancing humanoid robotics with backing from Hyundai
The competition mirrors broader trends in AI funding across Asia, where companies are quadrupling valuations as investors chase the next breakthrough technology.
| Company | Latest Valuation | Key Focus | Major Backers |
|---|---|---|---|
| Skild AI | $14+ billion | Robot-agnostic AI software | SoftBank, Nvidia, Samsung |
| Figure | $39 billion | Humanoid robots | OpenAI, Microsoft, Nvidia |
| Physical Intelligence | $5.6 billion | General-purpose robot brains | CapitalG, Thrive Capital |
| 1X | $10 billion (target) | Humanoid robots | OpenAI, Tiger Global |
Hardware Giants Seek AI Integration
The involvement of hardware manufacturers like Samsung and LG signals their recognition that software will differentiate future robotics products. These companies see Skild AI's platform as a way to accelerate their own robotics capabilities without developing AI from scratch.
Nvidia's participation aligns with its broader push into robotics computing. The chip giant has been investing heavily in AI startups whilst developing specialised hardware like the Jetson AGX Thor platform for robotics applications.
What makes Skild AI different from traditional robotics companies?
Skild AI focuses on creating universal software "brains" that can control any robot, rather than building physical robots. This robot-agnostic approach allows their AI to work across different manufacturers and machine types.
Why are Asian companies investing heavily in robotics AI?
Asian manufacturers like Samsung and LG see robotics AI as critical for maintaining competitiveness in automation and consumer products. They prefer partnering with AI specialists rather than developing capabilities internally.
How large is Skild AI's training dataset compared to competitors?
Skild AI claims its robotics dataset is 1,000 times larger than competitors, built through extensive simulations and millions of hours of internet video analysis.
What revenue has Skild AI generated so far?
The company grew from zero to $30 million in revenue in just a few months during 2025, demonstrating rapid commercial traction for its robotics AI platform.
Which industries could benefit most from robot-agnostic AI?
Manufacturing, logistics, healthcare, and domestic assistance are prime candidates. The universal nature of Skild's software allows deployment across multiple sectors without hardware-specific development.
The robotics AI arms race is accelerating with billions flowing into companies promising to democratise robot intelligence. As Asian enterprises struggle to move AI pilots into production, the success of companies like Skild AI will depend on bridging the gap between impressive demonstrations and reliable commercial deployment.
What do you think about the massive valuations in robotics AI, and which approach will ultimately succeed in the market? Drop your take in the comments below.
Latest Comments (6)
the article mentions Skild Brain's general-purpose model, which aligns with current research directions in multimodal foundation models for robotics. our team at RIKEN has been exploring similar architectures, and the progress seen in benchmarks like RoboNet and RLBench suggests that robot-agnostic approaches are indeed gaining traction.
It's interesting how Skild AI's focus on robot-agnostic foundation models parallels the trend we've observed with large language models. The transferability of AI is crucial for scalability, and this approach addresses critical concerns around responsible deployment across diverse physical systems, as discussed in IEEE's ethical AI frameworks.
seeing Skild AI valued at $14 billion now, that's wild. it shows just how much confidence is out there for foundational models, especially for robotics. makes you think about the potential for some of the incredible AI talent we've got brewing right here in Manchester and across the North. we're definitely playing in that league.
the idea of a robot-agnostic foundation model is really . it reminds me of how media platforms try to be device-agnostic now. skild brain doing dishes and stairs, okay, but the real test is how gracefully it handles cultural nuances in interaction. that's where the next level of "general purpose" comes in for me.
it's certainly interesting to see the investment flow into robot-agnostic foundation models, as mentioned with Skild AI. this parallels the trajectory we've observed in large language models, where pre-training on vast datasets allows for broad applicability. however, for robotics, the physical embodiment introduces complexities not present in text-based AI. while "Skild Brain" demonstrations are promising, the true test will be its performance on standardized benchmarks across various hardware platforms and environments, similar to how we evaluate LLMs on tasks like GLUE or SuperGLUE. robustness in real-world, unstructured settings remains a significant hurdle often downplayed in promotional materials.
I'm curious how "Skild Brain" will handle the sheer variety of tasks in a BPO setting. Like custom data entry or handling legacy systems, not just clearing dishes.
Leave a Comment