AI company Anthropic has launched Claude Sonnet 5.5, the latest addition to its Claude 5.5 family, with the company highlighting faster responses, improved coding capabilities and lower costs for several types of tasks.
Anthropic said Sonnet 5.5 generates responses more than 30 per cent faster than its predecessor, Claude Sonnet 5. The company also said the new model can cost up to 30 per cent less per task because it typically uses fewer tokens to complete the same work.
The model is aimed at coding, knowledge work, agent-based tasks and everyday applications. Anthropic said Sonnet 5.5 is particularly suited to software development tasks such as building features, fixing bugs, reviewing code and working with existing codebases.
On Anthropic’s FrontierCode evaluation, Sonnet 5.5 scored about 10 percentage points higher than Sonnet 5 at the same high-effort setting, while the company said the cost per task was roughly one-fifteenth that of the previous model. On CursorBench, which uses coding tasks based on real-world development workflows, Sonnet 5.5 recorded a 55.5 per cent score.
The company said the model also requires fewer tool calls and tokens on several coding tasks, allowing AI agents to complete work in fewer steps. Anthropic’s customer evaluations cited improvements in areas including software development, financial analysis and customer support.
Sonnet 5.5 retains the same standard API pricing as Sonnet 5, at $2 per million input tokens and $10 per million output tokens. Anthropic said its lower token usage can reduce the overall cost of completing tasks, while prompt caching and batch processing can provide additional savings.
The model is available through Claude’s web, iOS and Android applications, as well as Anthropic’s developer platform. It is also being offered through Amazon Web Services, Google Cloud and Microsoft Foundry.
Anthropic said Sonnet 5.5 includes additional cybersecurity safeguards and fallbacks developed for its more capable models, while routine software development and most life-sciences applications remain unaffected by those protections.
The launch follows the introduction of Claude Opus 5.5 earlier this month, with Anthropic positioning Sonnet 5.5 as a faster and lower-cost option for well-defined tasks, while Opus 5.5 is aimed at more complex work requiring extended reasoning and judgement.
