“Good morning. Andrew here. We’ve got news this morning on a new open-weight A.I. model that could challenge Anthropic and OpenAI — and it’s from an American start-up.
The implications of a successful U.S. open-weight model are significant. If companies move away from more costly closed models, the economics of the business models of Anthropic and OpenAI fundamentally weaken. That would batter their valuations, and could also hit those of other big tech giants, many of which have invested in the labs. More below. (Was this newsletter forwarded to you? Sign up here.)
A new open-weight A.I. contender
Anthropic is expected to go public next month in perhaps the biggest I.P.O. in history. Yet the artificial intelligence giant is facing a growing set of headwinds.
There’s the question of how safe its models are (more on that below). And then there’s increasing competition, not only from OpenAI, but also from cheaper open-weight models from China — and now from Reflection AI, an American start-up backed by Nvidia, Sri Muppidi reports.
Meet Beam, Reflection’s long-awaited model. Beam, introduced yesterday, is focused on coding and A.I. agents, applications popular with business users.
Reflection claimed that Beam’s reasoning performance is on par with GLM-5.2, a powerful open-weight model by the Chinese start-up Z.ai, and outperforms many Western open-weight ones.
More important, it requires less computing power to complete tasks than Chinese rivals, which is important for resource-hungry applications like agents, Stephanie Zhan, a partner at Sequoia Capital who invested in Reflection, told DealBook.
It’s the first volley in the A.I. race by Reflection, which was founded in 2024 by two former Google DeepMind researchers, Misha Laskin and Ioannis Antonoglou.
The company has raised billions of dollars from investors including Nvidia, Sequoia, Disruptive Technology Advisers and 1789 Capital, the investment firm at which Donald Trump Jr. is a partner. Reflection was most recently valued at $25 billion.
Open-weight models have become a big topic of debate in A.I. These tools, which users can modify, have risen in popularity as companies seek cheaper alternatives to closed-source models from places like Anthropic, Google and OpenAI.
Developers have increasingly adopted powerful open-weight models from Chinese labs, worrying leaders in Silicon Valley and in Washington. That has stoked interest in similar tools developed by Western companies like Meta; Thinking Machines Lab, which a former OpenAI executive founded; and Mistral, France’s A.I. champion.
Reflection adds to the challenges confronting Anthropic, as the company marches toward its I.P.O.:
Two of Anthropic’s biggest customers, Meta and Microsoft, are said to be trimming their use of the company’s Claude tools as a cost-saving measure, according to The Information. Even small reductions in use by major customers could dent Anthropic’s business. (In a draft I.P.O. prospectus obtained by Reuters, Anthropic disclosed that just two unnamed companies together accounted for roughly a quarter of its revenue last year.)
The Pentagon has stopped using Anthropic’s models amid a fight with the A.I. lab, according to BBC News.
Chinese labs are amassing not just users but also financial resources: DeepSeek is reportedly raising at least $12 billion in new funding and is weighing going public next year, as is Moonshot AI, the company behind the popular Kimi K3 model.” [1]
1. DealBook: Another challenger to Anthropic. New York Times (Online) New York Times Company. Oct 6, 2026.
Komentarų nėra:
Rašyti komentarą