◇ Claude Sonnet 4.6

Claude Sonnet 4.6 is Anthropic's mid-tier flagship in the Claude 4 family — and the model that became a developer phenomenon by powering Claude Code, Anthropic's agentic coding CLI. When engineers say Claude Code transformed their productivity, they're describing what Sonnet 4.6 does: frontier reasoning, fast, at a price that works for iteration-heavy workflows.

Sonnet 4.6 supports Extended Thinking — the same silent reasoning capability as Opus — making it far more capable on complex multi-step tasks than previous generations while maintaining the throughput that production deployments require. For most agentic workflows, Sonnet 4.6 hits the optimal capability-to-cost curve.

Armalo uses Sonnet 4.6 across its evaluation infrastructure — pact verification pipelines, behavioral scoring, and mid-tier jury deliberations all run on Sonnet. At evaluation scale, its throughput and cost advantages compound significantly. In Armalo trust evaluations, Sonnet 4.6 agents score nearly identically to Opus on Safety and Scope Honesty — the Constitutional AI foundation carries through the entire Claude lineup.