Anthropic has launched Claude Sonnet 5.5, a new mid-tier AI model designed for daily tasks such as document generation, coding, and data analysis. The company reports that the model operates over 30% faster than its predecessor and features improved writing clarity and human-like communication styles.
The pricing for the new model remains consistent with the previous generation, set at $2 per million input tokens, $10 per million output tokens, and $0.20 per million cache reads. Anthropic notes that the model's efficiency leads to a 30% reduction in overall costs for typical workloads.
Sonnet 5.5 includes invisible text watermarking to comply with international regulations, including the European Union’s AI Act. Additionally, it is the first model in the Sonnet series to incorporate cybersecurity safeguards and fallbacks previously reserved for more advanced models like the Mythos-class.
In performance evaluations, the model achieved a score of 70.6% on the Terminal-Bench 4.0 coding test, compared to 10.3% for the previous version. Zendesk reported that using the model helped resolve support tickets 20% faster, with fewer errors compared to previous production models.
Anthropic plans to expand its 5.5 model family by releasing Haiku 5.5 in the coming weeks. This upcoming version is intended to serve as the smallest and most efficient option within the Claude lineup, specifically optimized for high-volume and cost-sensitive applications.
