内容来源：https://aibusiness.com/foundation-models/reflection-ai-raises-2b-nvidia-leads-open-source-push

内容总结：

美国AI初创企业获巨额融资，开源模型赛道竞争加剧

由谷歌云赞助的生成式AI应用案例研究显示，当前行业正聚焦于提升人类信息交互体验。在这一背景下，两家美国人工智能初创公司近期引发市场关注。

总部位于纽约布鲁克林的开源AI企业Reflection AI宣布获得20亿美元新一轮融资，估值达80亿美元。该公司由两位前谷歌DeepMind研究员于去年创立，今年3月才以5.45亿美元估值完成1.3亿美元融资。本轮融资由英伟达领投，跟投方包括Disruptive、DST等风投机构，以及Zoom创始人袁征和前谷歌CEO埃里克·施密特。公司宣称已建成“能够训练大规模专家混合模型”的强化学习平台，并强调其开发“面向所有人的前沿开放智能”的定位，可能成为中美AI竞争中的美国代表力量。

与此同时，旧金山初创公司Worktrace AI正寻求以5000万美元估值募集1000万美元种子轮融资。该公司由前OpenAI产品经理和伊利诺伊大学助理教授联合创立，专注于通过AI识别和自动化重复性任务来优化企业工作流程。其投资方名单中出现了8VC、OpenAI基金等机构，以及前首席技术官米拉·穆拉蒂等多位OpenAI核心成员。

行业观察人士指出，这两家分别由谷歌DeepMind和OpenAI前成员创立的企业，反映出美国在开源AI模型领域的持续布局。前总统特朗普的AI顾问大卫·萨克斯公开表示，开源模型在成本优势和可定制性方面的特点，将吸引全球市场重要份额的关注。

中文翻译：

由谷歌云赞助
选定您的首批生成式AI应用场景
开启生成式AI之旅时，请首先关注能够优化人类信息交互体验的领域。

由前谷歌DeepMind研究员创立的布鲁克林初创公司打造开源大语言模型。前OpenAI员工领军的Worktrace AI寻求1000万美元种子轮融资。

自诩为“OpenAI开源替代者”的初创公司Reflection AI已确认获得20亿美元新一轮巨额投资，估值达80亿美元。这家总部位于纽约布鲁克林的企业通过官网及X平台发布声明，公布了最新融资成果。据Pitchbook数据显示，由谷歌DeepMind部门两位研究员于去年创立的Reflection AI于今年三月结束隐匿模式，当时以5.45亿美元估值筹集1.3亿美元。

英伟达主导本轮融资，其他参与方包括Disruptive、DST、1789、B Capital、Lightspeed、花旗集团、Zoom创始人袁征及前谷歌CEO埃里克·施密特。

Reflection在融资公告中强调，其正在构建“全民可及”的前沿开放智能体系，这或使其成为中美两国在AI领域竞争的重要参与者。公司声明称：“我们实现了曾被认为仅在全球顶尖实验室才能突破的创举——构建了能够训练前沿规模混合专家模型的大规模LLM与强化学习平台。在自动驾驶代码生成关键领域的成功实践，让我们亲眼见证了该方法卓有成效。突破这一里程碑后，我们正将此类技术拓展至通用智能体推理领域。”

Reflection对商业化扩张潜力充满信心，并获前总统特朗普任命的AI与加密货币事务负责人大卫·萨克斯公开称赞。萨克斯在X平台发文称：“见证更多美国开源AI模型崛起令人振奋。全球市场中有相当一部分客户青睐开源方案的成本优势、可定制性与自主控制权。我们期待美国在这一领域同样占据领先地位。”

与此同时，另一家由前OpenAI成员创立并获其支持的初创企业亦引发关注。

Worktrace AI由曾参与ChatGPT研发的前OpenAI产品经理安吉拉·江与伊利诺伊大学计算机科学助理教授迪帕克·瓦西斯共同创立，旨在通过AI识别并自动化重复性任务以优化企业工作流程。

该公司已获得8VC、Conviction、OpenAI基金、Genius Vultures及SV Angels等风投机构支持，同时吸引多位OpenAI核心人物投资，包括前首席技术官米拉·穆拉蒂、ChatGPT副总裁尼克·特利及OpenAI首席战略官杰森·权。

领英数据显示，总部位于旧金山的Worktrace现有2-10名员工。但据媒体报道，该公司正以5000万美元估值寻求1000万美元种子轮融资，引发其可能成为下一家获得重金注资的初创企业的市场猜测。

您可能还喜欢

英文来源：

Sponsored by Google Cloud
Choosing Your First Generative AI Use Cases
To get started with generative AI, first focus on areas that can improve human experiences with information.
Brooklyn startup founded by ex-Google DeepMind researchers builds open source LLMs. Worktrace AI, led by former OpenAI staff, seeks $10M seed funding.
Reflection AI, a start-up that sees itself as an open source alternative to OpenAI, has confirmed massive new investment of $2 billion at an $8 billion valuation.
Based in Brooklyn, New York, the company unveiled its latest funding success via a statement on its website and on X. Reflection AI, which was founded last year by two researchers from Google's DeepMind division, emerged from stealth in March, when it raised $130 million at a $545 million valuation, according to data from Pitchbook.
Nvidia led the latest funding round, with other participants including Disruptive, DST, 1789, B Capital, Lightspeed, Citi, Zoom founder Eric Yuan, and former Google CEO Eric Schmidt.
In announcing the funding, Reflection highlighted that it's building frontier open intelligence that is "accessible to all." This could position it as a U.S. rival to China's DeepSeek.
"We built something once thought possible only inside the world's top labs: a large-scale LLM and reinforcement learning platform capable of training massive Mixture-of-Experts (MoEs) models at frontier scale," the company said in the announcement. "We saw the effectiveness of our approach first-hand when we applied it to the critical domain of autonomous coding. With this milestone unlocked, we're now bringing these methods to general agentic reasoning."
Reflection said it is confident about the potential to scale commercially, with the company earning praise from President Donald Trump's so-called AI and crypto czar David Sacks, who posted on X: "It's great to see more American open source AI models. A meaningful segment of the global market will prefer the cost, customizability, and control that open source offers. We want the U.S. to win this category too."
Meanwhile another start-up -- founded and backed by former OpenAI personnel -- has also been attracting attention.
Worktrace AI is the brainchild of Angela Jiang, an ex-OpenAI product manager who worked on ChatGPT, and Deepak Vasisht, an assistant professor in computer science at the University of Illinois. It aims to streamline enterprise workflows by identifying and automating repetitive tasks via AI.
Worktrace already cites financial backing from venture capital firms such as 8VC, Conviction, the OpenAI Fund, Genius Vultures and SV Angels, plus a host of significant OpenAI figures such as former CTO Mira Murati; Nick Turley, vice president of ChatGPT; and OpenAI chief strategy officer, Jason Kwon.
LinkedIn currently lists Worktrace, which is based in San Francisco, as having between two and 10 employees, but according to media reports, it is looking to raise a seed round of funding of $10 million at a $50 million valuation, fueling speculation it could be the next start-up to attract heavy investment.
You May Also Like