Anthropic: Claude Sonnet 4.6

Anthropic's Claude Sonnet 4.6, created on Feb 17, 2026, is positioned as a highly capable Sonnet-class model excelling in iterative development, complex codebase navigation, and end-to-end project management. It demonstrates competitive response times, performing among the faster models with a 58th percentile speed ranking across 8 benchmarks. However, its pricing tends to be at premium levels, ranking in the 14th percentile for cost-effectiveness. A standout feature is its exceptional reliability, achieving a perfect 100% success rate across all benchmarks, indicating minimal technical failures. The model exhibits strong performance across various categories. It achieves perfect 100% accuracy in both General Knowledge and Ethics, with the latter also being the most accurate model at its price point and speed. Coding and Mathematics benchmarks show high accuracy at 94%, placing it in the 87th and 80th percentiles respectively. Hallucinations are well-managed with 98% accuracy, and Reasoning is robust at 88%. While Instruction Following is solid at 70%, it represents a relative area for potential improvement compared to its other high-performing categories. Email Classification is competent at 97%. Overall, Sonnet 4.6 is a highly reliable and accurate model, particularly strong in knowledge, ethics, and coding, albeit at a premium price point.