AMI Labs, a new venture co-founded by Turing Award winner Yann LeCun after leaving Meta, has raised $1.03 billion at a pre-money valuation of $3.5 billion. AMI is working on world models, or AI, that learn from reality, not just language.
There are fewer players in this category than generative AI, but probably not for long. “My prediction is that ‘world model’ will be the next buzzword,” AMI Labs CEO Alexandre LeBrun told TechCrunch. “Within six months, every company will be calling itself the world model for fundraising.”
LeBlanc said with a smile because he believes AMI Labs is fundamentally different. Its goal is to understand the real world. This could also have applications in the medical field, and AMI Labs’ first partner will be digital health startup Nabla, of which he is currently chairman.
As Nabla’s CEO, LeBrun had come to the same conclusion as LeCun about the limitations of large-scale language models (LLMs), where hallucinations can have life-threatening effects. But he also acknowledges that it will be some time before the startup offers a viable alternative based on LeCun’s proposed joint embedding prediction architecture, JEPA, in 2022.
“AMI Labs is a very ambitious project because it starts with basic research. It’s not your typical applied AI startup where you can release a product in three months, be profitable in six months, and make $10 million (annual recurring revenue) in 12 months,” LeBlanc said. In contrast, it can take many years for a world model to go from theory to commercial application.
Despite this period of time, companies developing world models are collecting hefty checks. SpAItial has raised a $13 million seed round, an unusual size for a European startup. Meanwhile, Fei-Fei Li’s World Lab secured a whopping $1 billion last month alone. Now, AMI Labs has joined the club with more funding than originally rumored.
The French AI institute reportedly sought just 500 million euros last December, but ended up raising around 890 million euros. Probably because of that team. In addition to Mr. LeCun’s involvement as chairman and Mr. Leblanc’s track record as an entrepreneur, Meta’s Vice President for Europe, Laurent Solly, will serve as Chief Operating Officer, renowned researcher Sainin Hsieh will serve as Chief Scientific Officer, Pascal Huang will serve as Chief Research and Innovation Officer, and Michael Labatt will serve as Vice President of World Models.
tech crunch event
San Francisco, California
|
October 13-15, 2026
According to LeBlanc, the high level of interest has given startups the opportunity to choose investors, both in terms of alignment of expectations and background. The round was co-led by Cathay Innovation, Greycroft, Hiro Capital, HV Capital, and Bezos Expeditions, with participation from several other funds and industry backers, as well as individuals including Tim and Rosemary Berners-Lee, Jim Breyer, Mark Cuban, Mark Leslie, Xavier Neal, and Eric Schmidt.
Value-add aside, this funding gives AMI Labs meaningful runway to fund two major cost centers: compute and talent. LeBlanc said he will prioritize quality over quantity as he builds out AMI Labs’ team across four major locations. New York, LeCun teaches at New York University. Montreal is the home of Rabat. And in Singapore, the aim is both to recruit AI talent and to get closer to future customers in Asia.
Although AMI Labs has no immediate plans to generate revenue, it still plans to engage with potential customers early. “We develop world models that try to understand the world, but you can’t do that in a lab. At some point, you have to put the models in a real-world situation with real data and real evaluations,” LeBlanc said.
When the time comes, AMI Labs will look to partners to explore implementation. Nabla is the first announced partner looking to gain access to these early models, but by no means the last. “This may explain the presence and strong interest of certain industry players and potential partners in the investment round,” LeBlanc said.
In addition to lead investors and angels, AMI Labs is backed by NVIDIA, Samsung, Sea, Temasek, Toyota Ventures, and French player associations Familiale Mulliez, Groupe Industriel Marcel Dassault, and Publicis Groupe. Aglaé Lab, Alpha Intelligence Capital, Artémis, Bpifrance Digital Venture, New Legacy Ventures, SBVA, and ZEBOX Ventures also participated.
It may take some time for these investments to translate into commercial applications. But staying true to LeCun’s beliefs, AMI Labs will continue to publish papers.
“We also plan to open source a lot of our code,” said LeBlanc, who previously worked at Meta’s AI research lab, FAIR. The startup’s founders still believe in open research, even though it’s “increasingly rare.” “We believe things move faster if we are open, and it is in our best interest to build a community and research ecosystem around us.”
