Anthropic has officially launched Sonnet 5.5, the latest addition to its Claude product line. This mid-tier model is designed for daily office tasks, including software development, document writing, and spreadsheet processing. The company positions it as a practical tool for enterprise users who prioritize operational efficiency over maximum capability.
The primary improvement in Sonnet 5.5 is its operational efficiency. Compared to the previous Sonnet 5 version released three months ago, the new model offers a 30% increase in output speed and reduced token consumption. These optimizations allow businesses to execute large-scale AI workflows at a lower cost.
In specific scenarios, such as agentic coding tasks, Sonnet 5.5 outperforms the higher-tier Opus 5.5 model. Anthropic attributes this to the model's ability to coordinate multiple agents simultaneously at a lower cost. Additionally, Sonnet 5.5 is the first in its series to be included in the company's advanced security management framework.
Anthropic also announced that an update for its lightweight Haiku model is currently in development. The new version of Haiku is expected to be released in the coming weeks, though a specific date has not been provided. This update follows a period of intense competition in the AI sector, with companies like OpenAI, Meta, and Google frequently updating their model lineups.
