Lovable: Gemini 4 Argon for App AI Features
LovableLovable apps can now use Google's Gemini 4 Argon (google/gemini-4-argon) for their AI features. The model accepts text, image, audio, and video input and is premium priced. Builders can ask for it by name or describe a goal and let Lovable choose.
Key Takeaways
- Lovable apps can now call
google/gemini-4-argonfor their built-in AI features. - The model accepts text, image, audio, and video input, covering the widest range of media Lovable offers.
- Google positions it for long-horizon coding, reasoning, and knowledge work.
- It is a premium-priced model, so builders should weigh cost against need.
- Builders can request it by name in chat.
- Alternatively, builders can describe their goal and let Lovable choose the model.
What Changed
Lovable added Google's Gemini 4 Argon to the models available for the AI features inside apps built on the platform. The model ID is google/gemini-4-argon. Google describes it as a frontier model for long-horizon coding, reasoning, and knowledge work.
Capabilities and Cost
Gemini 4 Argon accepts text, image, audio, and video as input, which makes it suited to apps that need to reason over several media types in one request. Lovable lists it as a premium-priced model, so it costs more than the standard options. The changelog does not state a specific amount.
How to Use It
Builders can ask Lovable to use Gemini 4 Argon directly, or describe what the AI feature needs to do and let Lovable pick a suitable model. This keeps model choice optional for people who do not want to compare options themselves.