feature
Lovable apps can now use Google's Gemini 4 Argon model for AI features
docs.lovable.devagent
Gemini 4 Argon is now available as a frontier model option for app AI features. It supports text, image, audio, and video input and is designed for long-horizon coding, reasoning, and knowledge work at a premium price. Developers can request it in prompts or let Lovable automatically select appropriate models.
Stacks this affects
Lovable ships in 1 tested stack here.
More Lovable
featureGPT Image 2.6 Flare and Sunburst bring image generation to LovablefeatureOpenAI's GPT Image 2.6 Flare and Sunburst models available for image generation and editingreleaseLovable adds GPT-6.1 Terra and new OpenAI image generation modelsfeatureTopaz Starlight Fast 2 model added for video restoration and upscalingfeatureThree new connectors: Fortnox, Mailchimp, and Mailchimp TransactionalreleaseLovable adds Topaz Starlight Fast 2 for video upscaling