At the start of the year, Chinese A.I. startup DeepSeek jolted Silicon Valley with a powerful open-source model that rivaled Western systems at a fraction of the cost. Out of that shockwave came Reflection AI, a New York-based startup founded by top A.I. researchers that has just raised $2 billion to build similarly accessible technology. The funding round valued the company at a staggering $8 billion, according to The New York Times.
Returning investors include Nvidia, Lightspeed Venture Partners and Sequoia Capital, while new participants include former Google CEO Eric Schmidt and 1789 Capital, the venture capital firm linked to Donald Trump Jr. Reflection previously raised about $130 million from backers like LinkedIn co-founder Reid Hoffman and Meta executive Alexandr Wang, according to Crunchbase, and was last valued at around $550 million in March.
Reflection AI emerged from stealth earlier this year, co-founded by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou. Laskin previously founded Claire AI, a startup focused on predicting retailer product demand. Antonoglou, DeepMind’s sixth-ever researcher, worked on the lab’s breakthrough AlphaGo model. Reflection AI is currently hiring for eight open roles across New York, San Francisco and London. Its growing staff includes former employees of OpenAI, Meta, Character.AI and Anthropic.
Like many of its rivals, Reflection aims to build superintelligence and is focused on autonomous coding tools. It recently released an A.I. coding agent called Asimov. Central to its mission, however, is the belief that open-source A.I. should be widely available to developers, researchers and individual users.
Reflection’s ambitious goal was spurred by the shock of DeepSeek’s emergence—an event many in the tech world likened to the Soviet Union’s 1957 launch of Sputnik 1, which spurred the U.S. space race. “The reason for our existence is we are living through a modern day Sputnik moment,” Laskin, who also serves as Reflection’s CEO, told CNBC SquawkBox today (Oct. 8). “When you think about what we’re competing against today, it’s the incredible open models that are coming from China.”
As A.I. becomes increasingly embedded in science, education, energy and supply chains, Reflection warns that concentrating such power in a few closed systems could stifle innovation and exclude other players. The company also argues that openness promotes safety, enabling a broader range of researchers—rather than a handful of private labs—to identify and address potential risks in A.I. systems.
Reflection is the latest A.I. model builder to benefit from surging investor enthusiasm for the field. According to Crunchbase, global venture capital funding during the third quarter of 2025 totaled $97 billion, representing a 38 percent increase compared to the same time period last year. Nearly half of that total, about 46 percent, went to A.I. firms, with the three largest rounds of the quarter raised by foundation model companies Anthropic, xAI and Mistral AI.
