Anthropic has made its new model, Claude Sonnet 4.5, available to all users. The model is considered by the company to outperform competitors like OpenAI and Google in software development, and it is being presented as “the world’s best code-writing model.”
Claude Sonnet 4.5 promises high-level performance in software programming (code generation), building complex AI agents, and computer usage capabilities.
This model is significant because it follows Anthropic’s previous models, Opus 4.1 and Sonnet 4. According to the product team, Sonnet 4.5 is smarter across all parameters while being more cost-efficient than Opus 4.1.
The AI giant emphasizes that Claude Sonnet 4.5 has taken the lead in industry-standard software tests. Additionally, the model can work autonomously, maintaining focus on multi-step tasks continuously for up to 30 hours.
The new model also shows strong performance in specialized fields such as cybersecurity, finance, and research. Claude Sonnet 4.5 will be automatically integrated as the default model on the Claude platform, while paid subscribers can choose to switch to Opus or previous Sonnet versions if desired.
Looking ahead, Anthropic has stated that by the end of 2025, it is very likely to release one or two additional versions, including a new Opus model.
