Anthropic has launched Claude Sonnet 5.5, the second model in its Claude 5.5 family, positioning it as a faster, more economical option for coding, everyday work, and document creation.

The model generates output more than 30% faster than its predecessor, Sonnet 5, and costs up to 30% less per task in the company's testing, according to Anthropic's announcement. The savings stem largely from efficiency: Sonnet 5.5 requires fewer tokens to complete the same work. Per-token pricing itself remains unchanged at $2 per million input tokens and $10 per million output tokens. That distinction matters for teams running the model at volume: because list prices are unchanged, the savings arrive at the task level, where costs accumulate with usage.

Anthropic is presenting Sonnet 5.5 as a lower-cost complement to Claude Opus 5.5. Opus remains aimed at complex, open-ended work that demands sustained judgment, while Sonnet is designed for more clearly defined tasks such as bug fixing and the creation of documents, slides, and spreadsheets.

On some evaluations, however, the gap between the two models has narrowed considerably. Sonnet 5.5 outscored Opus 5.5 on Terminal-Bench 4.0, recording 70.6% versus 66.4%. On GDPval-AA, a benchmark measuring real-world knowledge work, the two were nearly tied, with Sonnet 5.5 scoring 1,844 against 1,846 for Opus 5.5. Anthropic said that at certain effort settings Sonnet 5.5 performs on par with Opus 5.5, though the company continues to regard Opus as the stronger model for complex tasks requiring sustained judgment. For buyers weighing the two tiers, those results offer a reference point for how much measured capability is retained at Sonnet's lower per-task cost.

Coding is one of the largest areas of improvement. Anthropic reported that Sonnet 5.5 gained roughly 10 points over Sonnet 5 on FrontierCode at High effort while costing about one-fifteenth as much per task. The model's best CursorBench score also came within about two points of Opus 5.5.

Broader knowledge work also saw significant gains. Anthropic said Sonnet 5.5 came close to Opus 5.5 in computer use and chart recognition, and outperformed Sonnet 5 across the company's internal and external evaluations.

Sonnet 5.5 is also the first Sonnet model to ship with cybersecurity safeguards comparable to those on Anthropic's higher-end models, following a substantial improvement in its cyber capabilities over the previous generation. As a result, higher-risk cybersecurity requests may fall back to Sonnet 5, while routine development tasks remain available — a trade-off between tighter safeguards and task availability that teams using the model for security-adjacent work will need to account for.

The model is available across Anthropic's platforms as well as through Amazon Web Services, Google Cloud, and Microsoft Azure. Developers can access it through the Claude Platform using the identifier claude-sonnet-5-5.

Anthropic added that Claude Haiku 5.5, a model aimed at high-volume, cost-sensitive applications, will join the Claude 5.5 family in the coming weeks. Once it ships, the family is set to cover Anthropic's tier range end to end, from sustained complex work down to high-volume deployment.