Lovable Launches Aesthetics to Give Users More Control Over Design Output.

Lovable has launched Aesthetics, a new feature that gives users direct control over the visual design of their AI-generated web applications. The update allows builders to specify typography, layout, and color preferences during the chat-based development process, rather than relying solely on the tool's default design choices.

The feature, announced on LinkedIn by Lovable's Felix Haas, enables users to request multiple design concepts while building and produce what the company describes as "bolder" landing pages, apps, and blogs. The change addresses a common limitation of AI development tools: limited control over aesthetic decisions without diving into code.

Lovable's chat-first interface already generates full-stack applications from text prompts. Aesthetics extends that capability to the design layer, letting users iterate on visual style through conversation rather than manual CSS adjustments. A user could, for example, ask for a brutalist layout with monospace typography and a specific color palette, then request variations until the design matches their vision.

Aesthetics operates within Lovable's existing chat interface. Users can specify design preferences at project start or request changes mid-build. The feature surfaces multiple design concepts, which builders can compare and refine through follow-up prompts.

The update builds on Lovable's recent focus on user control. The platform recently shipped a new creative engine with multi-model routing and design previews, giving users more visibility into how the AI constructs their applications. Aesthetics takes that a step further by making design decisions explicit and negotiable.

For context, Lovable generates React and TypeScript applications using models like GPT-5.5 and Claude Opus 4.7. The platform handles deployment to Netlify or Vercel, manages authentication through Supabase, and recently added native payment support through Stripe and Paddle. Aesthetics slots into this stack as a design orchestration layer, translating high-level visual preferences into component-level styling decisions.

Most AI development tools prioritize speed over customization. They generate functional applications quickly but offer limited design flexibility without manual code edits. This works for prototypes but becomes a bottleneck for production apps that need brand consistency or specific user experience patterns.

Aesthetics attempts to solve this by making design a first-class citizen in the prompt-to-app workflow. Users can specify design constraints upfront, reducing the need for post-generation tweaks. This matters particularly for landing pages and marketing sites, where visual differentiation is critical.

Competing tools like v0 and Bolt generate UI components from prompts but typically require iteration through regeneration or manual editing. Lovable's approach surfaces design options during the build process, aiming to reduce the iteration cycle.

The feature also positions Lovable more directly against design-first tools like Framer and Webflow, which offer extensive visual control but steeper learning curves. Lovable's pitch is that Aesthetics delivers comparable design flexibility through natural language, without requiring users to learn a design interface.

Aesthetics is now available to all Lovable users. The feature does not appear to require additional subscription tiers or credits, though Lovable's standard token limits apply to design iteration requests.

For users building customer-facing applications, Aesthetics provides a practical way to differentiate their products visually without hiring a designer or learning CSS. The ability to request multiple concepts and iterate through conversation aligns with how non-technical founders often think about design: as a series of aesthetic decisions rather than technical implementations.

Lovable has not disclosed technical details about how Aesthetics translates user preferences into code. The feature likely combines prompt engineering, model fine-tuning, and component libraries to generate cohesive design systems from high-level descriptions. How well it handles edge cases and complex design requirements will determine whether it becomes a standard part of the Lovable workflow or remains a starting point for further customization.