Claude Sonnet 5 Comes to Lovable
Lovable now supports Claude Sonnet 5, which benchmarks show accomplishes more per task at lower cost, with stronger security and improved handling of risky requests.
Lovable has added support for Claude Sonnet 5, Anthropic's most agentic Sonnet model, bringing improved productivity, lower cost per task, and notably stronger security capabilities to the platform.
More Done for Less
According to Lovable's benchmarks, Claude Sonnet 5 gets more done per task while costing less than the previous Sonnet model. For creators building apps on Lovable, that translates directly into more output from each prompt — more features built, more iterations explored — without spending more credits.
Improved Security and Risky-Request Handling
A standout capability of Sonnet 5 in Lovable is its improved ability to catch and handle risky requests. The model is better at recognizing when a request could produce unintended or harmful output and responding appropriately — a meaningful upgrade for teams building production apps where reliability and safety matter.
Why It Matters
Lovable is built for creators who want to go from idea to working software without writing code. Sonnet 5's combination of stronger agentic reasoning, lower cost, and improved security makes it a more capable partner for that workflow — handling more of the complexity autonomously so you spend less time re-prompting and more time shipping.
Sources: the announcement on X.