Introduction
Anthropic has recently released Sonnet 5, a new AI model designed to offer near-Opus 4. 8 performance at a lower price point. In a blog post, Anthropic highlighted Sonnet 5 as a significant upgrade, especially for users who rely on Claude for coding, research, automation, and document work.
Sonnet 5 Performance
Sonnet 5 is built to be the most agentic Sonnet model yet, featuring advanced features such as the ability to make plans and use tools like browsers and terminals. Previously, these features were mostly locked to Opus 4. 8, but now Sonnet 5 can do almost everything the flagship model can.
Features and Capabilities
Sonnet 5 is designed to be more agentic, meaning it can handle tasks like creating plans or calling tools, and it's surprisingly good at verifying its own work. It handles queries like 'Can you fix your code? ' much better, almost as well as Opus 4.
Comparison with Opus 4.8
While Sonnet 5 is designed to be more agentic, it is not as expensive as Opus 4. The company offers introductory API pricing of $2 per million input tokens and $10 per million output tokens through August 31, 2026, with a higher rate of $3 per million input tokens and $15 per million output tokens after that period. This pricing structure makes Sonnet 5 a more cost-effective option for users who need to use more tokens than Sonnet 5, such as developers who use Opus 4.
User Experience
Users have described Sonnet 5 as 'much more agentic than its predecessors,' and it can check its own output without always being explicitly asked. This feature is particularly useful for developers who want to understand their own code and ensure its correctness.
Conclusion
Sonnet 5 is a significant upgrade for Anthropic's AI models, offering near-Opus 4. The company's pricing structure also makes Sonnet 5 a more attractive option for users who want to save money on API usage.
Comments