Anthropic releases Claude Sonnet 4.5 with powerful coding tools Apfelpatient

Anthropic has released Claude Sonnet 4.5, positioning it as the most powerful model for coding and logical reasoning. The company claims it outperforms GPT-5 and Gemini 2.5 Pro. In addition to significantly improved math and logic features, the model brings new tools for developers, increased application security, and broader integration into apps and browsers.

Competition in the field of AI models is fierce. OpenAI, Google, and Anthropic regularly release new versions of their systems, each one becoming increasingly more powerful. With Claude Sonnet 4.5, Anthropic is focusing on a combination of computing power, practical tools, and security. The goal is to provide a platform that takes coding, automation, and the use of agents to a new level.

Improvements in coding and logic

Claude Sonnet 4.5 is described by Anthropic as "the best coding model in the world." Compared to its predecessor, Opus 4.1, it demonstrates significant advances in logical reasoning and mathematics. This makes it not only a powerful tool for programming, but also for developing complex agents and using computers in automated processes.

New features in Claude Code

Claude Code has been enhanced with Claude Sonnet 4.5. New features include checkpoints, which save progress and allow you to revert to an earlier state at any time. This is a crucial advantage for long and complex projects. A new terminal interface has also been introduced, making code handling clearer and more flexible.

More options in the Claude apps

Claude's apps now support direct code execution and file creation. Tables, slides, or documents can be created directly within the conversation without having to open additional programs. Claude is now also available for Chrome. Initially, browser integration will be offered to select Max users during a test phase.

Claude Agent SDK for individual solutions

Another new feature is the Claude Agent SDK. This allows developers to create their own agents and flexibly adapt them to different requirements. Whether for businesses, research, or education – the SDK significantly expands Claude's possible applications.

Security as a focus

Anthropic has extensively trained Claude Sonnet 4.5 to reduce problematic behavior. The focus is on reducing flattery, deception, power-seeking, and avoiding hallucinatory content. The goal is to provide a model that is not only powerful but also reliable and trustworthy.

Anthropic sets clear accents with Claude Sonnet 4.5

Claude Sonnet 4.5 demonstrates Anthropic's strong influence on competition in the AI market. The model combines improvements in mathematics and logic with practical tools such as checkpoints, a terminal interface, file creation directly within apps, and Chrome integration. Added to this is the Claude Agent SDK, which enables customized solutions. The emphasis on security makes the model not only more robust but also more responsible in its use. Claude Sonnet 4.5 is now available worldwide, and Anthropic recommends upgrading to the new version. (Image: Shutterstock / PhotoGranary02)