Lovable Adds Claude Sonnet 5 for More Efficient, Secure Software Development.

Lovable has added support for Anthropic's Claude Sonnet 5, bringing what the company describes as more efficient development and stronger security capabilities to its chat-based software building platform.

The update, announced on LinkedIn, positions the new model as both faster and better at identifying risky requests compared to the previous Sonnet version. Lovable's internal benchmarks show users can accomplish more with less input, though the company did not publish specific performance metrics.

The core improvement centers on security. According to Lovable, Claude Sonnet 5 demonstrates enhanced capabilities for catching risky requests during the development process. This matters in a platform where users build software through natural language prompts: the model needs to understand not just what to build, but what might introduce vulnerabilities or expose sensitive data.

Lovable has been expanding its enterprise security posture in recent months. The company launched Workspace Insights in June, giving workspace admins visibility into every project being built across their organization. The addition of Claude Sonnet 5's improved risk detection adds another layer to this governance strategy.

The platform already supports multiple AI models. Users can switch between different versions of Claude and other supported models depending on their needs, balancing speed, cost, and capability. The Sonnet 5 addition gives teams another option when security considerations outweigh raw speed.

Lovable claims Claude Sonnet 5 delivers more output for less input, though the announcement did not quantify those gains. In practice, this likely means users need fewer follow-up prompts to refine generated code, or that the model makes fewer mistakes requiring correction.

The timing aligns with broader adoption trends. Lovable recently published research showing the emerging "build economy," where non-developers increasingly create software using AI-powered platforms. More efficient models lower the barrier further, reducing the iteration cycles that can frustrate non-technical builders.

The platform has been adding capabilities rapidly. Recent updates include native AI features without API keys, voice capabilities for speech-enabled apps, and a native Claude integration via MCP server. The Claude Sonnet 5 integration extends this momentum.

For Lovable users, the update is straightforward: select Claude Sonnet 5 from the model picker when starting or continuing a project. The change is most relevant for enterprise teams with security requirements, or individual builders working on applications that handle sensitive data.

The announcement reflects a broader shift in how AI coding tools compete. Raw capability matters less than the specific combination of speed, cost, security, and reliability that fits a given use case. By supporting multiple models including Claude Sonnet 5, Lovable lets users choose the right tradeoff for each project.

Lovable did not specify whether Claude Sonnet 5 support extends to all plan tiers or carries different pricing. The company also did not indicate whether this replaces or supplements the previous Sonnet model. Users can access the new model through the standard model selection interface in their projects.