Most agentic Sonnet to date; near-Opus quality at a fraction of the price.
Sonnet 5 is Anthropic's balanced tier and the workhorse many Claude developers actually build on. It is the most agentic Sonnet to date and delivers near-Opus quality at a fraction of the price, with a roughly 1M-token context that matches the flagship for long documents and codebases.
Pricing needs an asterisk: input is around $2 per million on introductory terms that may not persist, and output pricing has not been consistently published — so model the cost carefully before committing at volume. Sonnet is not the deepest reasoner in the lineup, and for the hardest problems Opus 5 remains the better tool.
For the broad middle of production work — agentic pipelines, coding, high-throughput tasks that still need real quality — Sonnet 5 is the pragmatic default, stepping up to Opus only where the extra reasoning earns its cost.
FIELD NOTE — The default workhorse for many Claude builders.
VISIT OFFICIAL SITE →