About Newsar
AI-powered news aggregation platform that shows you multiple perspectives on the same stories
Our Mission
In an era of increasing media polarization, Newsar aims to break down echo chambers by presenting news from multiple political perspectives. We believe that understanding different viewpoints is essential to forming well-informed opinions.
By using AI to automatically classify, cluster, and analyze news articles, we make it easy to see how the same story is covered across the political spectrum—from far left to far right.
What We Do
Multi-Source Aggregation
We collect news from dozens of sources across the political spectrum, giving you access to diverse perspectives in one place.
AI Classification
Using local AI models (Ollama), we automatically detect language, political bias, key entities, and sentiment for every article.
Story Clustering
Articles about the same events are automatically grouped together using vector embeddings and semantic similarity analysis.
Source Transparency
Every article shows its source's political bias rating, confidence scores, and classification method—full transparency, always.
Technology
Newsar is built with modern, open-source technologies that prioritize privacy and performance:
Frontend
- • Nuxt 4
- • Vue 3
- • Nuxt UI
- • UnoCSS
Backend
- • Nuxt Server API
- • PostgreSQL 16
- • Drizzle ORM
- • pgvector
AI/ML
- • Ollama
- • llama3.2:3b
- • nomic-embed-text
- • Local processing
Privacy First
All AI processing happens locally using Ollama—your reading habits and preferences never leave your device. We don't track you, sell your data, or use external AI services.
Newsar is designed to be self-hosted, giving you complete control over your news consumption and data privacy.
Questions or Feedback?
We're always looking to improve. If you have suggestions, found a bug, or just want to learn more about how Newsar works, we'd love to hear from you.