Nvidia-backed Reflection AI raises $2B at $8B valuation to build superintelligent open-source AI models

Reflection AI, the New York startup co-founded by former Google DeepMind researchers Misha Laskin and Ioannis Antonoglou, has raised $2 billion in new funding, valuing the company at $8 billion. Barely a year old, Reflection AI is already positioning itself as one of the most ambitious players in the open-source AI race.
This is one of the largest funding rounds ever for an open-source AI company, and it signals how quickly Reflection has gained the trust of investors. With fresh capital in hand, the startup is setting its sights on building large language models that can stand toe-to-toe with China’s DeepSeek, France’s Mistral AI, and U.S.-based Meta Platforms.
The round drew support from an impressive lineup, including former Google CEO Eric Schmidt, Citi, Donald Trump Jr.-backed private equity firm 1789 Capital, and existing investors Lightspeed and Sequoia. It comes just two months after The Information reported that Reflection was in talks to raise more than $1 billion.
Reflection was founded in 2024 with a clear focus: automating software development through AI. Its tools are aimed at transforming how code is generated and maintained, an area that has seen growing interest as AI shifts from experimentation to production environments.
With $2 Billion in Funding, AI Startup Reflection AI Aims to Take on DeepSeek, Meta, and Mistral in the Open-Source AI Race
The new funding marks a massive leap in valuation. Just months ago, Reflection raised $130 million at a $545 million valuation, according to PitchBook data. Now, with an $8 billion price tag and heavyweight investors behind it, the company is being positioned as one of the key challengers in the open-source AI ecosystem, going up against established players like OpenAI and DeepSeek.
“Reflection’s new valuation marks a sharp step up from its previous round, when it raised $130 million at a $545 million valuation, Reuters reported, citing PitchBook data.
Open-source AI has been moving at breakneck speed this year. The spark came in January with DeepSeek’s R1 model, which stunned the industry by matching GPT-4’s performance with a training budget of just $6 million. Released under an MIT license, R1 proved that top-tier models don’t have to rely on massive compute budgets.
Meta followed with its Llama series, which has racked up over 800 million downloads. Yann LeCun, Meta’s chief AI scientist, praised the collaborative nature of open-source work, saying DeepSeek’s team “came up with new ideas and built them on top of other people’s work. Because their work is published and open source, everyone can profit from it.”
Reflection is betting that blending efficiency with superintelligence will allow it to carve out a unique position in this fast-shifting landscape. In Europe, Mistral has built momentum in data-sensitive industries like banking and defense, raising more than $1 billion and reaching a $6 billion valuation. DeepSeek has been pushing price wars through ultra-efficient models. Meta has scale. Reflection wants to hit all three fronts at once: efficiency, precision, and global reach.
The open-source strategy comes with clear advantages—lower costs, transparency, and speed—but it’s not without risks. Cisco researchers recently revealed vulnerabilities in DeepSeek’s R1 that could be exploited through algorithmic jailbreaking. Any company looking to win enterprise adoption in this space will have to prove it can build systems that are both trustworthy and performant.
Much of Reflection’s new funding is already secured, and the company’s next phase will be judged by execution. To meet its ambitions, it will have to deliver models that match DeepSeek’s efficiency, Meta’s scale, and Mistral’s domain expertise—while staying true to the open-source ethos that’s fueling this new wave of AI development.
🚀 Want Your Story Featured?
Get in front of thousands of founders, investors, PE firms, tech executives, decision makers, and tech readers by submitting your story to TechStartups.com.
Get Featured