Outperformed GPT-4o in real-world tests on multi-file development tasks

Independent benchmarks validate Claude Sonnet 4.6 as a top-tier model for complex reasoning and coding tasks.

Showed only 3.7% self-preservation bias in a study testing AI deception, the lowest among prominent models tested.

Showed 77.7% self-preservation bias in a study testing AI deception, the highest among 23 models tested.

Used in prompt compression study analyzing 358 successful runs from 1,199 real orchestration instructions

Anthropic released Claude Sonnet 4.6 with native chain-of-thought reasoning mode for complex coding tasks

Service disruption with elevated error rates reported on status page