Claude Sonnet 5 Is Now the Default Across Every Claude Plan

Anthropic made Claude Sonnet 5 the default model across every Claude plan on June 30, 2026 — free, Pro, and Max tiers alike now route to it automatically rather than requiring users to manually select it. For a directory that lists Claude as one of its most-compared AI assistants, this is worth flagging on its own: the model most people encounter when they open Claude today is materially different from what was default just a few months earlier.

What's actually new

Early adoption reports point to Sonnet 5 as a clear step up specifically for coding, debugging, and multi-step agentic work — the kind of task where a model has to plan several actions ahead and keep track of state across them, rather than answer a single, self-contained question. Insurtech company Pace has reportedly put it into live production for multi-step insurance workflows, including intake and claims setup — a meaningful signal, since running an AI model on live business systems (rather than a demo or pilot) is a higher bar than most model announcements clear.

The pricing window

Anthropic introduced Sonnet 5 at introductory API pricing of $2 per million input tokens and $10 per million output tokens, running through August 31, 2026. If you're evaluating models for a new project, this is a reasonable window to test Sonnet 5 against whatever you're currently using, since the intro pricing won't last indefinitely.

One thing worth knowing if you build on the API

Sonnet 5 removed the temperature and top_p parameters that were available on earlier Claude models. If you have existing code that sets either of these, expect API errors until you remove them — a small thing, but the kind of detail that's easy to miss when upgrading and then confusing to debug after the fact.

Why this matters if you're choosing between assistants

Model updates like this are exactly why we track tools continuously rather than treating a listing as a one-time writeup — the Claude you'd be evaluating today isn't the same product it was two months ago. If coding and multi-step task automation are your primary use case, this is a good moment to actually re-test Claude against whatever you're currently using, rather than relying on an older impression.