release
Three new AI models available: GPT-6 Astra, Gemini 3.8 Flash, and Gemini 3.5 Transcribe
docs.lovable.devagent
Three new AI models are now available for apps: GPT-6 Astra for complex reasoning and coding, Gemini 3.8 Flash for multimodal chat, and Gemini 3.5 Transcribe for speech-to-text with 85+ language support. Developers can specify which model to use or let Lovable choose based on the task.
Stacks this affects
Lovable ships in 1 tested stack here.
More Lovable
featureLovable projects now sync with Bitbucket CloudfeatureConnect Lovable apps to Cloudflare APIfeatureProjects now support drafts for isolated testing before publishingfeatureDraft projects let you test changes in isolationreleaseLovable adds GPT-6 Astra, Gemini 3.8 Flash, and Gemini 3.5 Transcribe for app AI featuresfeatureLovable replaces message queue with follow-ups feature