Anthropic has unveiled its Claude 4 model

Anthropic has unveiled its Claude 4 model family, marking a significant advancement in AI capabilities for developers and enterprises. The flagship model, Claude Opus 4, is designed for complex coding and long-duration tasks, outperforming OpenAI’s GPT-4.1 with a 72.5% score on the SWE-bench benchmark. It can autonomously work for up to seven hours, maintaining consistent performance throughout. Alongside Opus 4, Anthropic introduced Claude Sonnet 4, a more accessible model that still offers enhanced reasoning and memory functions.

These models support "extended thinking with tool use," enabling them to switch between internal reasoning and external tools, such as web searches. They also feature improved memory capabilities, enabling better context retention over long tasks. Anthropic's focus on safety is evident, with Opus 4 classified under their highest risk category, ASL-3, due to its advanced capabilities.I

To facilitate integration, Anthropic has introduced the Model Context Protocol (MCP), an open standard that simplifies the connection of AI models to various data sources and tools. This protocol has gained adoption from major AI providers, including OpenAI and Google DeepMind.

Anthropic's developments position Claude 4 as a leading solution for AI-driven coding and intelligent agents, emphasising both performance and responsible deployment.