(Reuters) – Anthropic on Monday launched an advanced AI model that can produce faster responses or display its step-by-step reasoning process, as it looks to gain a competitive edge in the generative artificial intelligence industry.
The introduction of Anthropic’s hybrid model – which combines multiple reasoning approaches to solve complex problems more effectively – comes amid fierce competition in AI development, with U.S. tech firms vying against each other and Chinese companies such as DeepSeek and Alibaba.
The Amazon and Google-backed startup said the Claude 3.7 Sonnet model is its most advanced and will be available on all Claude plans, including Free, Pro, Team and Enterprise.
However, the “extended thinking mode” feature is only available on paid plans.
In extended thinking mode, the model “self-reflects before answering,” improving its performance on math, physics, instruction-following, coding, and many other tasks, Anthropic said.
The San Francisco-based company added that the hybrid reasoning model has been designed to focus on “real-world” tasks and less on math and computer science problems to reflect how businesses actually use large language models.
Anthropic said it is also releasing a limited-release preview of Claude Code, an agentic coding tool that helps developers with coding tasks, allowing them to “delegate substantial engineering work directly from their terminal.”
An agentic coding tool is an AI-powered software application that can autonomously perform coding-related tasks.
While users can choose how much time and resources are devoted to answering a question, the company said its pricing structure will remain the same as its previous models.
Anthropic’s new model is cheaper than rival OpenAI’s o1 model, costing $3 per million input tokens and $15 per million output tokens compared to $15 and $60, respectively.
(Reporting by Zaheer Kachwala in Bengaluru; Editing by Tasim Zahid)
Comments