Anthropic released Claude Sonnet 5.5 as the second model in its Claude 5.5 family, pitched as a faster, lowerâcost option for coding, everyday work and document creation. Anthropic reported Sonnet 5.5 is more than 30% faster than Sonnet 5 and can cost up to 30% less per task by requiring fewer tokens for the same work. Token pricing remains $2 per million input tokens and $10 per million output tokens. The model is available across Anthropicâs platforms and on AWS, Google Cloud and Microsoft Azure (claude-sonnet-5-5). Anthropic also said Claude Haiku 5.5 will arrive in the coming weeks.
Benchmarks show strengths and tradeoffs. On TerminalâBench 4.0 Anthropic reported Sonnet 5.5 at 70.6% versus Opus 5.5 at 66.4%; Artificial Analysisâs independent run found 63.6% for Sonnet 5.5 and 59.6% for Opus 5.5. On GDPvalâAA the models were effectively tied, 1,844 for Sonnet 5.5 and 1,846 for Opus 5.5. But test conditions diverge: Artificial Analysis noted Sonnet 5.5 at max effort wrote roughly 193,000 tokens per test task, which worked out to about $7.60 per task and shows higher token use than Opus 5.5. Anthropic emphasizes savings at lower effort settings (Medium), while acknowledging Opus 5.5 remains preferable for complex, sustainedâjudgment work. Sonnet 5.5 also launched with improved cybersecurity safeguards compared with prior Sonnet models.