Anthropic has unveiled Claude Sonnet 5.5, its mid-tier model. According to the company, the new version runs 30 percent faster than its predecessor, consumes fewer tokens, and outperforms Opus 5.5 on agent-based coding tasks.
Anthropic has officially announced Claude Sonnet 5.5, an updated version of its mid-tier Sonnet model. According to the company, the new model runs faster than its predecessor and can be used at a lower cost.
Arriving roughly three months after the release of Sonnet 5, this version is positioned as an ideal assistant for everyday tasks and the creation of office documents.
The previous model, Sonnet 5, stood out for its efficient deployment of AI agents and its lower-cost operation compared with competitors. According to Anthropic, Sonnet 5.5 takes these advantages a step further by offering a directly speed-focused improvement.
Based on the company's figures, the new model runs 30 percent faster than its predecessor and significantly reduces token consumption as well.
Positioned below Opus, the most powerful tier in Anthropic's model hierarchy, Sonnet can be more advantageous in certain scenarios thanks to the agility it offers. The benchmarking tests shared by the company show that Sonnet 5.5 outperforms Opus 5.5 in agent-based coding tasks. Behind this result is the ability to deploy multiple AI agents simultaneously without exceeding cost limits.
Sonnet 5.5 also draws attention with its cybersecurity capabilities. Anthropic says the new model has cybersecurity abilities comparable to Opus 5. With this development, Sonnet 5.5 becomes the first Sonnet model to be subject to the strict cybersecurity reviews and safeguards applied to the Fable and Opus models.
The company also announced that it plans to release a new version of its smallest model, Haiku, in the coming weeks, though it did not share a specific date.
Log in or sign up to leave a comment.
No comments yet. Be the first!
