Nvidia-Backed Startup Seeking to Counter Chinese AI Eyes $25 Billion Valuation
Reflection is one of several startups working alongside Nvidia to build powerful, freely available ‘open-source’ AI models
Reflection, a startup backed by chip giant Nvidia NVDA 1.99%increase; green up pointing triangle that is leading an effort to create freely available U.S. AI systems, is in talks to raise $2.5 billion at a valuation of $25 billion, according to people familiar with the matter.
The company is one of a handful of Nvidia-linked startups that are seeking to build a network of “open source” AI models, which businesses, labs and universities can use and repurpose according to their needs.
The deal would add firepower to Reflection, which is central to Nvidia’s efforts to create an open-source ecosystem that can run on its chips and counter the growing and increasingly sophisticated offerings in China. A spokesperson for Reflection declined to comment.
JPMorgan Chase is in talks to participate in the round through its newly formed Security and Resiliency Initiative, which was created in December to back U.S. companies in industries critical to the economy and national security, the people said. The bank said at the time that it planned to invest up to $10 billion into venture-backed startups through the initiative.
Disruptive, an earlier investor in Reflection, is also expected to invest.
Reflection has developed close ties with Nvidia, which invested about $800 million in a previous funding round that valued the young startup, which has yet to generate meaningful revenue, at $8 billion, the Journal has reported.
The $25 billion valuation for Reflection represents its value before the $2.5 billion investment, the people said.
The deal is part of a flurry of investments in so-called neolabs, which are giving priority to long-term research and building new AI models over short-term profits.
Nvidia has been introducing Reflection to potential customers. That includes foreign governments that plan to develop homegrown AI infrastructure that gives them power over how the technology is used.
Reflection recently committed several billion dollars to build models customized for the Korean language in a deal with South Korean conglomerate Shinsegae Group. Thousands of Nvidia chips will power the data center supporting the project. Reflection plans to make more deals like this one, with the goal of becoming the open model used by sovereign clouds in U.S. allies around the world, the people familiar with the matter said. AI boosters have taken to calling such systems “sovereign AI.”
Investors have described Reflection as the “DeepSeek of the West”—an alternative to the open-source models offered by Chinese companies. U.S. AI labs haven’t given priority to open-source models, creating concerns that other countries will end up using Chinese technology instead.
“Open models are Trojan horses for the infrastructure they bring with them,” Reflection Chief Executive Misha Laskin said in an interview earlier this month.
Laskin, a Russian-Israeli researcher who previously worked for Google’s DeepMind, started the company in 2024. It has raised more than $2 billion to date, not including the latest financing, from investors including 1789 Capital, where Donald Trump Jr. is a partner, as well as Lightspeed Venture Partners, CRV and Sequoia Capital.
Reflection is also part of a new consortium of startups, called the Nemotron Coalition, that Nvidia has assembled to develop open AI models. Other members include AI startups Cursor and Thinking Machines Lab.
The Financial Times earlier reported some details of the investment.