Anthropic launched Claude Sonnet 5.5 on September 28, pitching a faster mid-tier model whose claimed cost advantage weakens under some high-effort tests.
The Anthropic launch post lists token rates of $2 per million input tokens and $10 per million output tokens, unchanged from Sonnet 5. Anthropic claims the new model runs more than 30% faster and costs up to 30% less per task. Its published Terminal-Bench 4.0 score is 70.6%, ahead of Opus 5.5 at 66.4%. But Artificial Analysis reported that a maximum-effort run produced about 193,000 output tokens and cost roughly $7.60 per task—around 50% more than Sonnet 5 in that setup.
Claude Sonnet 5.5 is available through the Claude Platform, AWS, Google Cloud and Microsoft Azure. Developers migrating agents that run Sonnet with thinking disabled must adopt the new between_tools setting.
Teams should benchmark complete tasks at the effort levels their agents actually use, rather than treating lower token prices as a guaranteed lower bill.