Anthropic released Claude Sonnet 5.5 on September 28 8. Anthropic describes it as a clear upgrade over Sonnet 5 that runs 30%+ faster and costs up to 30% less for most work 1. TechCrunch reported the same speed claim, attributed to Anthropic 7.
Where the saving comes from
The list price did not move: $2 per million input tokens and $10 per million output tokens, the same as Sonnet 5 2. Anthropic says the saving comes from the model needing fewer tokens to finish the same job, up to 30% less per task in its own testing 3.
Then on October 7, alongside the Haiku 5.5 launch, Anthropic halved the price of Sonnet 5.5 cache reads 6. If your app reuses the same long instructions on every call, that change matters more than the list price.
One setting to know about
In Claude Code and the Claude apps, the default effort is Medium. On the API platform, it defaults to High 4. Lower effort answers faster and uses fewer tokens, so if a hard task comes back thin in the app, raising effort is the first thing to try.
What Anthropic says it is not for
Anthropic is direct that Opus 5.5 remains clearly stronger at complex, open-ended work that needs sustained judgment 5. Sonnet is the everyday worker, not the strategist.