release
Three new AI models available for Lovable app features: GPT-6 Astra, Gemini 3.8 Flash, and Gemini 3.5 Transcribe
docs.lovable.devagent
Lovable apps can now use OpenAI's GPT-6 Astra for complex reasoning and coding, Google's Gemini 3.8 Flash as the new default chat model supporting multimodal input (text, image, audio, video), and Gemini 3.5 Transcribe for automatic speech-to-text. This significantly expands the AI model choices available to developers building applications.
Stacks this affects
Lovable ships in 1 tested stack here.
More Lovable
featureLovable adds Parallel connector for web search and researchfeatureLovable integrates Parallel for web search and researchfeatureLovable introduces free Chat feature for workspace collaborationfeatureLovable adds AI chat assistant for workspace and projectsfeatureLovable adds Jev Latest model for typed decision-making in AI featuresreleaseLovable API released for workspace automation and management