Summary

  • Anthropic has released its latest generation of AI models, optimised for problem-solving and coding.
  • Claude Opus 4 is the company’s most powerful model to date, and can handle long-running tasks for several hours, according to customer tests.
  • It outperformed rivals, including Google, OpenAI and GPT-4.1, on coding tasks and using web search tools.
  • Claude Sonnet 4 is a more affordable model, suited to general tasks and said to be more precise.
  • Both incorporate ‘thinking summaries’, and an ‘extended thinking’ beta feature will allow users to switch between modes.
  • Anthropic’s Claude Command Line Tool is now widely available, following a limited preview in February.
  • The company plans to issue more frequent model updates to keep pace with competitors.

By Jess Weatherbed

Original Article