Anthropic Unveils Sonnet 4.5: The Safest AI Model Yet
In a significant breakthrough, Anthropic has introduced Sonnet 4.5, the company's latest artificial intelligence (AI) model that outperforms its predecessors and competitors in coding and agentic tasks. According to Anthropic, Sonnet 4.5 is not only the world's best coding model but also the safest AI system developed by the company to date.
Sonnet 4.5 has set a record score of 61.4 percent in OSWorld, a suite that tests AI models on real-world computer tasks, surpassing its predecessor Opus 4.1 by 17 percentage points. The new model can autonomously work on multi-step projects for more than 30 hours, a significant improvement from the seven or so hours Opus 4 could maintain at launch.
"We're thrilled to introduce Sonnet 4.5, which represents a major milestone in our efforts to build agentic systems that can perform complex tasks independently," said Dario Amodei, CEO of Anthropic. "Our team has worked tirelessly to ensure that this model is not only highly capable but also safe and reliable."
Anthropic's emphasis on safety is crucial, as the development of advanced AI models raises concerns about their potential misuse. Sonnet 4.5 has undergone extensive testing and evaluation to mitigate these risks.
The introduction of Sonnet 4.5 comes less than six months after Anthropic announced Opus 4 and Sonnet 4, two new AI systems that marked a significant improvement over the company's previous models. However, Sonnet 4.5 outperforms not only its predecessors but also more expensive competitors like Google's Gemini 2.5 Pro and GPT-5 from OpenAI.
The development of agentic AI systems like Sonnet 4.5 has far-reaching implications for society. As these models become increasingly capable of performing complex tasks independently, they may revolutionize industries such as healthcare, finance, and education. However, experts warn that the risks associated with advanced AI must be carefully managed to ensure their safe deployment.
Anthropic's focus on safety is a welcome development in the AI community, where concerns about model reliability and accountability have grown in recent years. As Sonnet 4.5 continues to push the boundaries of what is possible with AI, it remains to be seen how this technology will shape the future of human society.
In related news, Anthropic has announced plans to release more information about its safety protocols and testing procedures for Sonnet 4.5 in the coming weeks. The company's commitment to transparency and accountability is a crucial step towards building trust in AI research and development.
As the AI landscape continues to evolve at breakneck speed, one thing is clear: Anthropic's Sonnet 4.5 represents a significant leap forward in the development of safe and reliable AI models.
*Reporting by Engadget.*
