Japan’s SoftBank Group and Nvidia are in advanced talks to invest in Skild AI in a funding round worth more than $1 billion, a move that could value the maker of foundation models for robots at about $14 billion, according to sources and a term sheet reviewed by Reuters.
If the deal closes on schedule before Christmas, it would nearly triple the company’s valuation from the $4.7 billion it reached in a $500 million Series B round earlier this year, which drew in Nvidia, LG’s venture capital arm, and Samsung, among others, according to PitchBook data.
Founded in 2023 by former Meta AI researchers, Skild AI has emerged as one of the fastest-growing players in the bid to solve one of robotics’ hardest challenges: building general software that can serve as the decision-making core for machines of multiple shapes and uses. Instead of manufacturing hardware, Skild develops foundation models trained on vast datasets so robots can perceive their environment and make decisions with human-like fluidity. The ambition is to overcome a major constraint that has kept general-purpose robots from operating widely in both homes and industrial settings.
Register for Tekedia Mini-MBA edition 19 (Feb 9 – May 2, 2026): big discounts for early bird.
Tekedia AI in Business Masterclass opens registrations.
Join Tekedia Capital Syndicate and co-invest in great global startups.
Register for Tekedia AI Lab: From Technical Design to Deployment (next edition begins Jan 24 2026).
The company is backed by Amazon.com and Lightspeed Venture Partners, adding to the momentum around firms building the “brains” for the next generation of robots. Skild AI raised $300 million at a $1.5 billion valuation in its Series A round last year, with investments from Jeff Bezos, SoftBank, and Khosla Ventures.
The latest talks highlight the surge of investor interest in humanoid and general-purpose robotic systems as advances in artificial intelligence make these machines more capable of handling intricate tasks once considered far beyond reach. Heavyweights such as Nvidia, Samsung, AMD, and a rising group of specialized robotics software firms are locked in a race to supply the computational engines and learning systems that will underpin the category.
Even so, experts say the world is still several years away from seeing general-purpose robots deployed at scale. The technical constraints remain steep, including real-time decision-making, fine-grained manipulation, long-duration safety, and the cost of integrating such models into commercially viable machines. The industry is advancing quickly, but not at the pace implied by some of the recent enthusiasm.
SoftBank and Skild AI did not immediately respond to Reuters when asked for comment, while Nvidia declined to comment on the matter. A source familiar with the negotiations said some details could still change because the talks are fluid, though the goal is to complete the deal before Christmas. Another person close to the discussions said SoftBank was particularly impressed by Skild’s technology during pilot projects, which reinforced the company’s belief that software-driven robotics could define the next chapter of automation.
Robotics has become a centerpiece of SoftBank CEO Masayoshi Son’s strategic plans. In October, SoftBank acquired the robotics business of Swiss engineering group ABB in a $5.4 billion deal, adding to the portfolio that Son wants to build as part of a long-term robotics and automation push. The investment discussion around Skild AI fits that larger objective of gaining pole position in what Son expects to be the next major technology wave.
The funding effort comes at a moment when the U.S. government is also accelerating its attention on robotics. Commerce Secretary Howard Lutnick is holding meetings with industry CEOs to speed up development, and the Trump administration is weighing an executive order on robotics next year, according to Politico last week.
Policymakers see robotics as a strategic industry tied to productivity, supply-chain resilience, and national competitiveness, especially as manufacturing transitions toward more automated production architectures.
Skild AI introduced its first general-purpose AI model in July, presenting it as a system that can handle a wide variety of environments and tasks ranging from warehouse logistics to household chores. The company argues that robots powered by such models could become far more versatile, reducing the need for custom engineering approaches that have slowed adoption in many industries.
Investor excitement around Skild mirrors the broader surge in robotics funding this year as companies, governments, and chipmakers position themselves ahead of what they believe will be a transformative decade. Several countries are racing to establish dominance in humanoid robotics, and China recently warned that speculative activity in its own robotics sector had reached levels that risk creating a bubble, with more than 150 companies vying to launch humanoid robots.
The competitive backdrop underscores why companies like SoftBank and Nvidia are moving quickly. Nvidia’s role is especially notable because the company has become the central supplier of the chips and compute infrastructure that power modern AI systems. Its interest in Skild fits its broader campaign to lock in the software and developer ecosystems that will rely on Nvidia hardware as humanoid systems evolve. For SoftBank, Skild represents an opportunity to strengthen its robotics strategy with software that can scale across multiple robot manufacturers and use cases.
If the round closes at the expected valuation, Skild AI would enter 2026 as one of the most valuable robotics software companies in the world, with backing from some of the most influential names in technology and venture capital. The challenge, as always, will be delivering on the promise of general-purpose robotics.