In an era of fast-paced information that often induces stress, news-sentimental-analytic serves as a news curation solution centered on the mental health of the Indonesian public. By leveraging the natural language processing (NLP) capabilities of the Gemma model, this project processes local news articles to map text sentiment polarity with high accuracy. The service empowers users to filter news based on its emotional impact, aiming to reduce the doomscrolling cycle and build a healthier information ecosystem so citizens can stay informed without sacrificing their peace of mind.