Anthropic Launches Sonnet 5.5, Offering 30% Speed Improvement Over Sonnet 5

Anthropic has released Sonnet 5.5, which is 30% faster than its predecessor, Sonnet 5. This model is also the first in the Sonnet series to implement cyber safeguards, aligning it with the protections already in place for the Opus and Fable models. Notably, Anthropic claims that Sonnet 5.5 outperforms Opus 5.5 in agentic coding tasks and possesses comparable cyber capabilities to Opus. The release follows the announcement of Sonnet 5 just three months ago and comes shortly after OpenAI's new model releases last week. Additionally, Anthropic plans to unveil a new version of its smallest model, Haiku, in the coming weeks. This development positions Sonnet 5.5 as a significant player in the mid-tier AI model market, competing directly with OpenAI's Sol and Luna models.