Antigravity has unveiled the Gemini 3.5 Flash (Low) model, aimed at reducing token consumption by approximately 45% for simple tasks compared to the Gemini 3.5 Flash (Medium). This move comes in response to user feedback regarding excessive token usage. The new model also surpasses the performance of the previous Gemini 3 Flash (High) in software engineering tasks. To support developers, Antigravity has reset the Gemini quota limits across all plans for the upcoming week.