By Adam Levine
At its Code w/Claude event on Thursday, AI start-up Anthropic announced its new group of AI language models, Claude 4.0 Sonnet and Claude 4.0 Opus. Started by ex-OpenAI employees, Anthropic's AI models are widely used by software developers in particular.
Like many AI companies, Anthropic had struggled to make the next generation of large language models, and Claude 3.5 Opus, announced about a year ago, never materialized and was eventually scrapped.
Opus 4.0, therefore, could be a milestone. The name indicates it's the top performing and most expensive model in the Claude 4.0 series.
Model-makers have worked around the technical problems by creating "reasoning" models that work through problems in a drawn-out step-by-step fashion. This produces better answers, but can also take several minutes to get a response.
Model-makers hope that a new generation of models can produce the same quality of responses as the reasoning models, with much greater speed.
In brief testing by Barron's, Claude 4.0 provided high-quality answers much faster than OpenAI's current top model, ChatGPT o3.
The announcement of Claude 4.0 Opus would suggest that Anthropic is the first company to break through this technical ceiling. Others like OpenAI and Alphabet's Google could soon follow.
Write to Adam Levine at adam.levine@barrons.com
This content was created by Barron's, which is operated by Dow Jones & Co. Barron's is published independently from Dow Jones Newswires and The Wall Street Journal.
(END) Dow Jones Newswires
May 22, 2025 13:22 ET (17:22 GMT)
Copyright (c) 2025 Dow Jones & Company, Inc.
免責聲明：投資有風險，本文並非投資建議，以上內容不應被視為任何金融產品的購買或出售要約、建議或邀請，作者或其他用戶的任何相關討論、評論或帖子也不應被視為此類內容。本文僅供一般參考，不考慮您的個人投資目標、財務狀況或需求。TTM對信息的準確性和完整性不承擔任何責任或保證，投資者應自行研究並在投資前尋求專業建議。