Anthropic Introduces Claude Sonnet 5.5
Anthropic's new Sonnet model is faster, more efficient, and built for coding and everyday knowledge work.
Anthropic has introduced Claude Sonnet 5.5, the second model in its Claude 5.5 family. The company says the new model runs more than 30% faster than Sonnet 5 and can cost up to 30% less per task because it uses fewer tokens to complete the same work.
The launch comes after Anthropic introduced Claude Opus 5.5 earlier this month. Anthropic positions Sonnet 5.5 as a faster, lower-cost complement to Opus 5.5, with a focus on well-scoped everyday tasks, coding, and creating documents, slides, and spreadsheets.
Claude Sonnet 5.5 Brings a Large Coding Upgrade
Coding is one of the biggest areas of change. Anthropic reports that Claude Sonnet 5.5 scores 70.6% on Terminal-Bench 4.0, compared with 10.3% for Sonnet 5. On CursorBench 4.0, it scores 55.5%, compared with 34.1% for the previous model.
Anthropic says Sonnet 5.5 is also more efficient during coding tasks. Early testers reported that the model can batch tool calls more often, reducing the number of steps needed to complete work and lowering the overall token cost.
The release builds directly on the earlier Claude Sonnet 5 launch, which brought stronger agentic coding, tool use, and knowledge-work capabilities to the Sonnet line.
Faster Work Beyond Coding
Anthropic is also targeting broader knowledge work with Claude Sonnet 5.5. The company reports improvements in document creation, presentations, spreadsheets, computer use, image understanding, and long-running tasks. It says the model is designed to make fast iteration easier on work that does not require the deeper judgment associated with Opus 5.5.
That puts the model in the middle of Anthropic's growing Claude lineup. The company's recent Claude Opus 5.5 launch introduced its higher-end 5.5 model, while Claude Sonnet 5 vs. Opus 4.8 coverage provides background on how Anthropic's model tiers have been positioned.
The broader lineup has also expanded through Claude Fable 5.1 and Mythos 5.1, alongside the earlier Claude Opus 5 release. Those releases show how quickly Anthropic is building separate models around different levels of capability, access, and workload requirements.
Pricing and Availability
Claude Sonnet 5.5 keeps the same token pricing as Sonnet 5: $2 per million input tokens and $10 per million output tokens, with cache reads priced at $0.20 per million tokens. Anthropic says the model typically costs less per task because it needs fewer tokens to complete the work.
Claude Sonnet 5.5 is available now across Anthropic's platforms, including Amazon Web Services, Google Cloud, and Microsoft Azure. Developers can access it through the Claude Platform using the model identifier claude-sonnet-5-5.
Sonnet 5.5 Expands Anthropic's Claude Push
The new model arrives as Anthropic continues expanding Claude beyond a single chatbot experience. Recent Saganote coverage has followed Claude's voice mode and connected actions, Claude Tag for Slack, and the transition from Claude Cowork to Claude.
Anthropic is also pushing Claude into more specialized work, including Claude Science and research that led to a novel enzyme system discovered with Claude. Claude Sonnet 5.5 adds another general-purpose model to that expanding ecosystem, with speed and efficiency positioned as key parts of the release.