Sources familiar with the matter say that robot AI software startup Physical Intelligence is in discussions for a new funding round, aiming to raise at least $1 billion.
This round would nearly double the company's valuation to roughly $20 billion.
The people said existing investors Avenir and Lux Capital are in talks to lead the round.
As AI technology advances, building systems that can understand and adapt to the physical world has become feasible, driving a surge in investor interest in the robotics sector.
One source noted that Physical Intelligence completed a funding round earlier this year at a valuation of $11.5 billion.
Despite the soaring valuation, the company currently has no revenue, the sources said.
The startup has also drawn backing from Alphabet (GOOGL.US)'s independent growth fund CapitalG, Amazon founder Jeff Bezos, and Thrive Capital.
Founded in 2024, Physical Intelligence's founding team includes former Google DeepMind researchers as well as academics from Stanford University and UC Berkeley.
The company is developing AI software that is said to run on a variety of third-party hardware systems to perform various robotic actions.
In recent job postings, the company said employees will train robots to complete a series of common household chores, such as picking up groceries and placing them into bags, sorting cups and plates into bins, and installing light bulbs.
The startup faces growing competition from an increasing number of startups and large companies, including AI leader OpenAI and Elon Musk's Tesla.
Fellow startup Skild AI, which focuses on helping robots learn, reached a valuation of $14 billion in a January funding round and announced in September that its annual recurring revenue had surpassed $100 million.
Earlier this month, Physical Intelligence opened access to its latest robot model for external developers to access and fine-tune.
The service, still in testing, has generated strong interest from investors, according to the sources.
In its announcement, the company said: "Real-world robotics applications are not easy. Whether it's robotic construction, robotic surgery, or warehouse automation, it requires a whole-company effort."
Comments