Anthropic has launched its advanced AI model, Claude 3.7 Sonnet, which offers faster response times and step-by-step reasoning. The hybrid model combines multiple approaches to tackle complex problems more effectively. Available on all plans, including the free version, it aims to compete in the generative artificial intelligence industry.
The company’s focus is on “real-world” tasks, rather than math and computer science problems. A new feature, extended thinking mode, provides self-reflection before answering, improving performance on various tasks. Anthropic has also released a limited preview of Claude Code, an agentic coding tool that automates coding tasks. Pricing remains the same as previous models, with a competitive cost compared to rival OpenAI’s o1 model.
Note: The article is rewritten to retain essential information while simplifying the content, and adhering to the 10-word limit for the title.
Source: https://edition.cnn.com/2025/02/24/tech/anthropic-ai-hybrid-reasoning-model-hnk-intl/index.html