NewsBlur
FreemiumOpen-source RSS reader with intelligence training that learns what stories you like, social sharing, and the ability to read original sites within the app.
What does this tool do?
NewsBlur is an open-source RSS reader that combines traditional feed aggregation with machine learning-powered content filtering. It learns your reading preferences through explicit training (marking stories as liked or disliked) and surfaces relevant stories accordingly. The platform offers native applications across web, iOS, and Android, with unique features like reading original article content inline, full-text search across subscriptions, story tagging, and a social sharing layer called 'blurblog.' It also includes modern conveniences like daily briefing summaries, AI-powered question answering about stories, and the ability to follow non-RSS sources like YouTube channels. The intelligence training system differentiates it from basic RSS readers by actively filtering noise rather than simply listing all new items.
AI analysis from Feb 23, 2026
Key Features
- Machine learning-powered story filtering that learns from user training (likes/dislikes) to surface relevant content
- Full article text extraction from truncated RSS feeds with original site layout rendering
- Real-time push notifications for feed updates across web and mobile platforms
- Full-text search across all subscriptions with saved search feeds and advanced filtering by date range
- Story tagging system with searchable tags and custom organization
- Public and private blurblog sharing to curate and share stories with audience
- AI-powered question answering about story content and natural language search/filtering
- Daily briefing summaries customizable by feed or folder
Use Cases
- 1Individual journalists and researchers who need to monitor dozens of industry publications and apply custom filters to surface story angles
- 2Knowledge workers tracking news across multiple sectors (tech, finance, science) with intelligent filtering to reduce information overload
- 3Community managers curating content to share with followers through public blurblog feeds based on trained preferences
- 4Self-hosted deployments for organizations wanting complete data ownership and control over feed processing
- 5Email newsletter subscribers consolidating newsletters into a unified reader instead of managing inbox clutter
Pros & Cons
Advantages
- Intelligent training system genuinely learns preferences over time and actively reduces noise, unlike passive RSS readers that just push all stories
- Cross-platform native apps (web, iOS, Android) with feature parity means no functionality gaps between platforms
- Open-source with self-hosting option via Docker provides complete data ownership and customization capabilities
- Extracts full article text from truncated feeds and displays original site layouts, solving a major friction point in RSS reading
- Real-time feed updates and archive-everything model with searchable history across all subscription history
Limitations
- Free tier limited to 64 feeds is restrictive for power users; premium pricing ($5-$25/month) needed for meaningful feed counts
- Learning curve for training system—users must actively mark stories as liked/disliked for intelligence to become effective; passive readers won't benefit
- AI features (Ask AI, natural language filters) appear limited to paid tiers, fragmenting the feature set
- Self-hosting option requires Docker/technical knowledge; not accessible to non-technical users wanting self-hosted deployment
- No mention of collaborative features or team workspaces; designed for individual readers rather than teams
Pricing Details
NewsBlur offers a free tier supporting up to 64 feeds with real-time updates and basic training. Premium ($5/month) allows 1,024 feeds with 5x more frequent updates, river of news, and searchable tags. Premium Archive (higher tier, price not specified) adds 4,096 feed support, full archive with searching, global intelligence training, and AI features. Pro tier (highest, price not specified) supports 10,000 feeds with regex training and priority support. All new accounts include a 30-day free premium trial without requiring a credit card. Self-hosted option available with Docker for complete data ownership.
Who is this for?
Information professionals and power users who consume content from 50+ sources and want intelligent filtering rather than drowning in feeds. Target includes journalists, researchers, analysts, content curators, and knowledge workers in tech/finance/science. Also appeals to privacy-conscious users wanting self-hosted options. Less suitable for casual readers or teams needing collaboration features.