Reflection AI是一家自称美国对开源中国人工智能的回应的初创公司，正在发布其第一款名为Beam的模型。
Reflection AI, the startup billing itself as America’s answer to open-source Chinese AI, is releasing its first model, called Beam.
该公司由前DeepMind研究人员于2024年创立，将Beam定位为表现优于其他西方开放模型的“主力”。联合创始人兼首席执行官米莎·拉斯金（Misha Laskin）告诉Semafor，该模型在编码和代理任务方面尤其强大。他说，它还需要比同类开放模型少三到四倍的计算能力来推理问题，这使得它运行起来更快、更便宜。
The company, founded in 2024 by former DeepMind researchers, is positioning Beam as a “workhorse” that performs better than other Western open models. The model is particularly strong at coding and agentic tasks, cofounder and CEO Misha Laskin told Semafor. It also needs three to four times less computing power to reason through a problem than comparable open models, he said, making it faster and cheaper to run.
该型号的性能与中国Z.ai最近推出的型号之一GLM-5.2相当; Beam的性能正在“接近”阿里巴巴最先进的型号Qwen 3.8-Max，落后于OpenAI和Anthropic的顶级封闭美国型号。
The model’s capabilities are comparable to one of China’s Z.ai’s recent models, GLM-5.2, which was released in June; Beam is “approaching” the performance of Alibaba’s most advanced model, Qwen 3.8-Max, which trails the top closed American models from OpenAI and Anthropic.
中国的初创公司和科技巨头在开源人工智能方面处于世界领先地位。几家大型西方公司最近采用了中国模式，作为降低某些重复性任务（例如客户服务和编写常规代码）成本的一种方式。北京还将开源作为帮助其他国家技术发展的外交工具。但由于开放模型可以在更少的保护措施下下载和运行，因此它们的兴起也引发了安全担忧：美国和英国政府评估人员发现最近的中国模型可能会帮助黑客利用代码漏洞。
China’s startups and tech giants have led the world in open-source AI. Several large Western companies have recently embraced Chinese models as a way to reduce costs for certain repetitive tasks like customer service and writing routine code. Beijing has also promoted open source as a diplomatic tool to help other countries in their tech development. But because open models can be downloaded and run with fewer safeguards, their rise has also raised security concerns: US and UK government evaluators found that recent Chinese models could help hackers exploit code vulnerabilities.
拉斯金表示，总部位于纽约的Reflection认为，企业、政府和其他实体拥有一个巨大的市场，这些实体正在构建主权人工智能系统，这些系统不或不能使用中国模式，但仍然希望“拥有”和控制他们使用人工智能的方式。“他们今天真的没有很好的选择，”他说。
New York-based Reflection sees a large market in businesses, governments, and other entities building sovereign AI systems that don’t or can’t use Chinese models, but still want to “own” and control how they use AI, Laskin said. “They don’t really have very good options today,” he said.
Reflection已经在训练其下一个模型，拉斯金表示该模型将比Beam“更强大”、更有性能。
Reflection is already training its next model, which Laskin said would be “much more” powerful and performant than Beam.
这家初创公司获得了英伟达、红杉和花旗集团的投资，该初创公司于6月确认以250亿美元的资金前估值结束了最新一轮融资。随着Beam的发布，Reflection加入了其他西方初创公司的行列，包括Thinking Machines Lab和Mistral，这些公司也提供开放重量人工智能模型。
The startup, which has investments from Nvidia, Sequoia, and Citigroup, in June confirmed it closed its latest funding round at a $25 billion pre-money valuation. With Beam’s release, Reflection joins a slate of other Western startups, including Thinking Machines Lab and Mistral, that also offer open-weight AI models.
在发生一系列令人震惊的安全事件后，美国对人工智能监管的呼声达到了白热化的程度。在白宫会议后，美国领先的人工智能实验室和科技公司签署了美国总统唐纳德·特朗普所说的一项关于安全原则的“道德约束力”协议。
Its release comes as calls for AI regulation reach a fever pitch in the US, after a slate of alarming security incidents. The leading US AI labs and tech companies signed what US President Donald Trump called a “morally binding” agreement on safety principles following a White House meeting.
特朗普一直在推动公司自我监管，而不是让它们接受联邦监督，但如果民主党在11月接管国会，政治算计可能会迅速发生变化。
Trump has pushed for companies to regulate themselves, rather than subject them to federal oversight, but the political calculus could shift quickly should the Democrats take over Congress in November.
拉斯金说，反思组织希望“开放模型的构建者能够为华盛顿正在进行的监管讨论做出贡献”。
Reflection says it wants to enable “builders of open models to be able to contribute to the discourse that is happening in DC on regulation,” Laskin said.
他说：“你希望谈判桌上有多种声音，无论是公开的还是封闭的，他们应该合作帮助政府监管人工智能。”“我们认为这个入口可以使更开放的模型观点成为华盛顿话语的一部分。”拉斯金表示，Reflection正在与美国超级智能创新和标准推进中心和英国人工智能安全研究所合作，评估该模型的能力。
“You want multiple voices around the table, both open and closed, and they should be working collaboratively on helping the government regulate AI,” he said. “We see this entrance as something that enables more open model perspectives to be part of the discourse in DC.”Laskin said Reflection is working with both the US Center for Advancing Innovation and Standards for Super Intelligence and the UK’s AI Safety Institute to assess the model’s capabilities.