Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute cost
Reflection debuts Beam, an open-weight AI model to rival Chinese models at lower compute cost
Reflection 发布 Beam：一款旨在以更低算力成本挑战中国模型的开源权重 AI 模型
Reflection AI is officially unveiling Beam, its first frontier, open-weight AI model. The two-year-old, Brooklyn-based startup claims Beam matches the performance of leading Chinese open models on advanced reasoning benchmarks at dramatically lower costs, a claim that could heat up the race to build a Western answer to DeepSeek, Qwen, and Z.ai.
Reflection AI 正式发布了其首款前沿开源权重 AI 模型 Beam。这家位于布鲁克林、成立两年的初创公司声称，Beam 在高级推理基准测试中表现与中国领先的开源模型相当，但成本却大幅降低。这一声明可能会加剧西方构建对标 DeepSeek、通义千问（Qwen）和 Z.ai 等模型产品的竞争。
Reflection’s announcement confirms reporting from Axios over the weekend that the startup was close to a launch. The company shared new details in a lengthy blog post Monday, which described Beam as a text-only mixture-of-experts model trained on high-compute reinforcement learning to be effective at reasoning, coding, and agentic tasks at “a fraction of the token cost and inference time compute” of rivals.
Reflection 的公告证实了 Axios 周末的报道，即该公司即将发布产品。周一，该公司在一篇详尽的博客文章中分享了更多细节，将 Beam 描述为一种纯文本的混合专家（MoE）模型。该模型通过高算力强化学习训练，在推理、编码和智能体任务方面表现出色，且“Token 成本和推理时间算力仅为竞争对手的一小部分”。
Beam is a 501-billion-parameter model with 23 billion active parameters. It was pretrained on 23.8 trillion tokens and has a 1 million token context window. To compare, Z.ai’s GLM-5.2 has roughly 744 billion total parameters with 40 billion active.
Beam 是一个拥有 5010 亿总参数、230 亿激活参数的模型。它在 23.8 万亿个 Token 上进行了预训练，并拥有 100 万个 Token 的上下文窗口。作为对比，Z.ai 的 GLM-5.2 总参数约为 7440 亿，激活参数为 400 亿。
Reflection’s performance claims haven’t been independently verified, but on advanced reasoning benchmarks, the company says Beam scores on par with Z.ai’s GLM-5.2 and outperforms today’s leading Western open models while using “3-4x less inference compute.” Reflection calls it a “workhorse model” for enterprises, the public sector, and developers.
Reflection 的性能声明尚未经过独立验证，但该公司表示，在高级推理基准测试中，Beam 的得分与 Z.ai 的 GLM-5.2 持平，并优于目前西方领先的开源模型，同时推理算力消耗减少了“3 到 4 倍”。Reflection 将其称为面向企业、公共部门和开发者的“主力模型”。
Reflection is positioning itself against closed labs like Anthropic and OpenAI, against popular open models from Chinese developers, and against Western players like Mistral, Meta, and Cohere. Its most direct U.S. rival might be Inkling, the open model from Mira Murati’s Thinking Machines Lab released in July. Reflection’s own benchmarks show that Beam outscores Inkling on four coding tests where both report results, but Inkling is a multimodal model and Beam is text-only.
Reflection 的定位是与 Anthropic 和 OpenAI 等闭源实验室、中国开发者的流行开源模型，以及 Mistral、Meta 和 Cohere 等西方厂商竞争。其在美国最直接的竞争对手可能是 Mira Murati 的 Thinking Machines Lab 于 7 月发布的开源模型 Inkling。Reflection 自己的基准测试显示，在双方均有结果的四项编码测试中，Beam 的得分超过了 Inkling，但 Inkling 是多模态模型，而 Beam 仅支持文本。
Reflection was founded in 2024 by two former Google DeepMind researchers and has raised roughly $4.7 billion from backers, including Nvidia, Sequoia Capital, and Lightspeed Venture Partners, per PitchBook. Its last round valued the company at a $25 billion pre-money valuation.
据 PitchBook 数据显示，Reflection 由两名前 Google DeepMind 研究员于 2024 年创立，已从英伟达、红杉资本和光速创投等支持者处筹集了约 47 亿美元。其上一轮融资前的估值为 250 亿美元。
The startup has also been locking up compute — a key ingredient needed to train frontier models capable of luring customers away from Anthropic’s and OpenAI’s closed models, as well as the cheaper open-weight models from Chinese labs. This summer, Reflection signed deals collectively worth more than $7 billion with SpaceX and Nebius to secure access to Nvidia’s GB300 chips through 2029.
这家初创公司一直在锁定算力——这是训练前沿模型的关键要素，旨在吸引客户从 Anthropic 和 OpenAI 的闭源模型，以及中国实验室更便宜的开源权重模型转向其产品。今年夏天，Reflection 与 SpaceX 和 Nebius 签署了总价值超过 70 亿美元的协议，以确保在 2029 年前获得英伟达 GB300 芯片的使用权。
Reflection is aiming Beam and future models at enterprises and sovereign nations. The pitch is to build “AI factories,” a product that would let institutions build their own customized, local AI system by training Reflection’s AI models on their own proprietary data. Nvidia CEO Jensen Huang, whose company backs Reflection, has long championed the “AI factory” idea and pushed to strengthen the open AI ecosystem — a vision that would also benefit Nvidia, whose GPUs would power those systems. Axios reported that hedge funds and trading firms are among those eager to build such systems. Reflection has already begun testing the concept of a sovereign AI factory partnership with Shinsegae Group in South Korea.
Reflection 将 Beam 及未来模型的目标定位于企业和主权国家。其核心理念是构建“AI 工厂”，该产品允许机构通过在自有专有数据上训练 Reflection 的 AI 模型，来构建定制化的本地 AI 系统。英伟达首席执行官黄仁勋（其公司是 Reflection 的支持者）长期以来一直倡导“AI 工厂”理念，并推动加强开源 AI 生态系统——这一愿景也将使英伟达受益，因为其 GPU 将为这些系统提供动力。Axios 报道称，对冲基金和贸易公司是渴望构建此类系统的群体之一。Reflection 已经开始与韩国新世界集团（Shinsegae Group）测试主权 AI 工厂的合作概念。
Reflection says it will release Beam’s weights and full technical details this month, with distribution through hyperscalers and neoclouds and integrations across open source libraries at launch. Reflection did not respond in time to TechCrunch’s requests for more information.
Reflection 表示，将于本月发布 Beam 的权重和完整技术细节，并在发布时通过超大规模云服务商和新兴云平台进行分发，并集成到各大开源库中。截至发稿，Reflection 未能及时回复 TechCrunch 的置评请求。