Anthropic presented the Claude Sonnet 4.5 model with extended programming capabilities

Vlad Cherevko I have been interested in all kinds of electronics and technologies since the early 2000s. I like to play computer games and understand the work of different gadgets. I regularly monitor the news of the technology in the world and write materials about it.

Anthropic has announced the launch of a new Claude Sonnet 4.5 artificial intelligence model, which is positioned as the most productive in programming. According to the developers, the model is able to create not only prototypes, but also “ready for production” of applications, which is a significant step forward in the reliability of the code compared to previous models of artificial intelligence.

Claude Sonnet 4.5 is available through API and Chat Bot Claude. The cost for the developers remained at the previous version of Sonnet 4: $ 3 per million incoming tokens and $ 15 for a million weekends.

According to Techcrunch in the last year, anthropic models have been actively used in the field of software development. Apple and Meta are known to use them in internal processes. In addition, the API Claude is integrated into Cursor, Windsurf and Replit. At the same time, the competition in the market increased after GPT-5 exit from Openai, which exceeded the previous anthropic models in a number of tests.

The company states that Claude Sonnet 4.5 demonstrates the best results in several benchmarks, including SWE-Bench Verified. Anthropic researcher David Hershi reported that during the tests, the model could work for more than 30 hours autonomously, creating an application, adjusting the databases, buying domain names and even conducting an audit SOC 2 to check the product safety.

Anthropic emphasizes that the model has become more coordinated in the answers, as well as reduced the level of enthusiasm and erroneous statements. In addition, protection against attacks due to “fast injection” has been improved.

Together with CLAUDE SONNET 4.5, the company has introduced Claude Agent SDK – a set of tools to create its own agents based on the same infrastructure used in Claude Code. In addition, MAX subscribers have a preview of the Imagine With Claude function that demonstrates real -time software generation.

The Claude Sonnet 4.5 release took place less than two months after the Claude Opus 4.1 release, which indicates a high rate of renewal of artificial intelligence models.