Anthropic launched a new advanced AI model on Monday, designed to offer quicker responses or show its reasoning process step-by-step. This release aims to give the company a competitive edge in the fast-growing generative AI market, where US tech firms are competing alongside Chinese companies like DeepSeek and Alibaba.
The new model, called Claude 3.7 Sonnet, combines multiple reasoning techniques to solve complex problems more effectively. It will be available across all Claude plans—Free, Pro, Team, and Enterprise—but the “extended thinking mode” feature is only available on paid plans.
In extended thinking mode, the model “self-reflects” before providing an answer, which enhances its performance on tasks like math, physics, coding, and following instructions. Anthropic emphasized that the hybrid reasoning model is designed to focus more on “real-world” tasks, reflecting how businesses typically use large language models, rather than just math and computer science problems.
Additionally, the company is rolling out a limited preview of Claude Code, an agentic coding tool aimed at developers. This tool helps with coding tasks and allows developers to delegate significant engineering work directly from their terminal.
Despite these new features, Anthropic’s pricing structure remains the same as with previous models. The Claude 3.7 Sonnet model is priced lower than OpenAI’s o1 model, with a cost of $3 per million input tokens and $15 per million output tokens, compared to OpenAI’s $15 and $60, respectively.
