Drowning in information but starved for genuinely important news? Samuel Clay felt the same way, and that frustration on a New York City subway ride sparked the creation of NewsBlur. He wasn't aiming to launch a company, but to build a better way to filter the overwhelming flood of online content. NewsBlur's story began with a simple need: to cut through the noise and focus on what truly matters.
NewsBlur launched publicly in mid-2010, following an initial beta phase during which Clay used his developer blog to share updates with the small community of early adopters. By then, the core idea was already clear. RSS readers existed, but most of them did one thing badly: they showed you everything your subscriptions had published, in chronological order, with no sense of what you actually wanted to read. Clay wanted something smarter. He wanted a reader that learned.
The company was formally incorporated in 2009, though the work continued as a side project until 2012, when Clay transitioned to full-time development. Reforgers' profile on NewsBlur notes the company received backing from Y Combinator in the summer of 2012, a milestone that gave Clay the runway to stop freelancing and focus entirely on the reader he had been building in the margins of everything else.
March 2013: When an Earthquake Built a Mountain
For the first three years of its existence, NewsBlur grew slowly and steadily a small tool for people who cared about RSS, maintained by one developer who believed the protocol still had a future. Then Google killed Google Reader.
In March 2013, Google announced it would shut down its popular news reader service. The RSS ecosystem had grown deeply dependent on Google's free, well-indexed tool. Bloggers, journalists, researchers, and obsessive news consumers had built their daily workflows around it. When it disappeared, the vacuum was enormous and NewsBlur was there to fill it.
The numbers moved fast. According to the Wikipedia entry on NewsBlur, the platform's subscriber base rose from approximately 1,500 users to over 60,000 almost immediately following the announcement. That kind of growth would have broken most startups. For a one-person operation running on a side project budget, it was a crucible.
Clay wrote about the experience on the NewsBlur Blog, documenting what he called "Three Months to Scale NewsBlur" a detailed account of the infrastructure scramble, the late nights, and the lessons learned from trying to serve a user base that had grown by 4,000% in a matter of weeks. The post is a useful artifact of what it looks like when a small tool meets a suddenly massive audience, and what it costs to keep it running.
Intelligence Training: Teaching the Feed to Listen
What distinguished NewsBlur from the beginning and what continues to distinguish it now was not just that it could aggregate feeds, but that it could be trained. The feature is called Intelligence Training, and the idea is deceptively simple: instead of relying on an algorithm to infer what you want to read based on your click history, you tell the reader directly. Thumbs up an author. Thumbs down a tag. Train on a title keyword to highlight it in green, or fade away stories that don't interest you.
"No algorithm decides for you you're in control," reads the NewsBlur Features page. That sentence sounds like marketing language, but in this context it describes something the platform has genuinely built into its architecture. The training you do on one feed stays on that feed. The training you do across all feeds shapes your global River of News a combined scrollable stream that surfaces the best stories from across your subscriptions based on your specific preferences.
Power users can go further. Premium Pro subscribers can write regular expressions regex classifiers that match against story titles, body text, and URLs. The use cases are specific and practical: catching CVE numbers as they appear, tracking stock ticker patterns, identifying specific phrase combinations across multiple feeds and authors. For professionals who need to monitor news in real time, this is not a consumer toy. It is a precision instrument.
What the Platform Does in 2026
The NewsBlur of 2026 is a more mature product than the one that survived the Google Reader exodus. It has native apps for iOS and Android, not wrapper versions of a web app but first-class native clients built for each platform. It supports offline reading. It has dark mode. It integrates with IFTTT and Zapier, allowing users to build automated workflows around the stories they save and the feeds they monitor.
The platform can follow any website, even ones that do not publish an RSS feed. NewsBlur generates a custom feed for any web page by monitoring it for changes. Users can follow Twitter profiles, Reddit threads, GitHub repositories, product pages, and job boards. If it exists on the web, NewsBlur can watch it.
Story clustering groups duplicate stories across feeds so if the same news breaks across multiple outlets, readers see a consolidated view more than five variations of the same headline. Full-text extraction pulls complete articles from feeds that truncate their RSS output, displaying the original story in its entirety without requiring a click-through. Every story is kept in the archive and searchable forever on Premium Archive and Pro plans.
AI features have been added carefully. Ask AI lets users ask questions about any story and receive answers powered by artificial intelligence. Daily Briefing delivers a customizable summary of your top stories every morning. These features appear on the Premium Archive tier and above, reflecting a measured approach to AI integration that augments more than replaces the reader's own curation work.
The Four Rivers: How NewsBlur Decides What to Show
In July 2026, Clay introduced two new curated feeds to the NewsBlur sidebar: a redesigned Global Shared Stories and a new feed called Good Reads. These join Widely Read Stories and Long Reads, which launched in April 2026. Together they form what the NewsBlur Blog describes as "four curated rivers sitting together in the sidebar, each answering a different question."
The design philosophy is worth dwelling on, because it reveals something about how NewsBlur thinks about quality alongside engagement. None of the four ranks by clicks. None of them is trying to keep you scrolling. They are built out of what NewsBlur readers actually did with their time.
Global Shared Stories asks: what did people choose to hand to someone else? It runs on sharing, a deliberate human act. The rebuilt version, as Clay described on the blog, gathers every story shared across the site in the last few hours, caps each person at three shares so no single active user can flood the pool, drops private blurblogs, and then uses Claude Haiku to rank a few worth reading. The AI does not go looking for stories. It does not write anything. It only ranks stories that readers already chose to share, and it weighs most heavily the comments attached to those shares because a share with a few sentences is a share somebody thought about.
Widely Read Stories asks what held the most attention across NewsBlur. It runs on reading time, not clicks. A headline nobody read cannot buy its way in. Long Reads asks what was worth an afternoon features and essays that readers gave real time to more than skimmed. Good Reads asks what somebody finished and then kept, and it deliberately tilts toward feeds with few subscribers. "The big sites do not need help getting seen," Clay wrote. A story from a site with forty readers can beat a story from a site with forty thousand, if two people read it closely for thirty seconds or more and at least one of them saved, shared, or trained it up.
The Business Model: Sustainable Independence
NewsBlur operates on a freemium model. Limited access is free for up to 64 sites. Unlimited access requires an annual subscription. The features page outlines the current pricing structure: Premium at $36 per year (1,024 sites, search, text view, River of News), Premium Archive at $99 per year (full archive, Ask AI, daily briefing, 4,096 sites), and Premium Pro at $29 per month (5-minute feed fetching, regex training, 10,000 sites).
The open-source version of NewsBlur is published under the MIT License and is available on GitHub. This is not a hobby project that abandoned its open-source roots it is a commercial product that chose to give its core software away, maintaining community contributions and transparency around its architecture. The MIT license is permissive: anyone can run their own instance, fork the code, or build on top of it.
The Reforgers profile lists the company as active, with a team size of one. That one person is Clay, who has been running NewsBlur as a sustainable independent business since 2012, surviving the Google Reader exodus, building native apps, adding AI features, and maintaining the open-source version all without venture-scale funding or a large team. The sustainability is deliberate. The product does not need to be everything to everyone. It needs to be exactly what its users rely on it to be.
Why This Matters
For readers researching how independent software projects survive in an era dominated by platform consolidation, NewsBlur offers a specific case study. It did not pivot to social media. It did not sell user data. It did not bundle itself into a suite of productivity tools or rebrand as an "AI-powered knowledge management platform." It remained what it always was: a reader, focused on the act of reading, and on giving the reader more control over what appears on their screen.
The RSS ecosystem contracted sharply after Google's exit. Feedly, Inoreader, and other alternatives gained market share. But NewsBlur never tried to compete on scale. It competed on depth the Intelligence Training feature, the story clustering, the curated rivers built from human behavior signals more than engagement metrics. For a certain kind of reader, these are not nice-to-haves. They are the entire point.
The four new feeds launched in 2026 represent an evolution in thinking about what a news reader can surface. Most platforms optimize for recency or popularity or personalization based on inferred interest. NewsBlur's approach is different: it asks what readers did with their time, and it uses that data not clicks, not shares alone, but time spent and deliberate action to surface stories that were worth the attention they received.
Where to Read Further
To explore NewsBlur's origin story and the technical challenges Clay faced scaling from 1,500 to 60,000 users almost overnight, the NewsBlur Blog remains the primary record. Clay's "Three Months to Scale NewsBlur" post documents the infrastructure scramble in detail. For a comprehensive feature overview, the features page covers every current capability, from Intelligence Training to Ask AI to the four curated rivers.
Those interested in the platform's history and its place in the RSS ecosystem can consult the Wikipedia article for a structured overview, and the Grokipedia page for additional historical context on the founding years and early development. The Reforgers profile provides a business-level summary of the company's Y Combinator backing and current operational status.
For readers who want to try the platform itself, the free tier allows access to up to 64 sites a sufficient sample to test Intelligence Training and experience the reading interface before committing to a paid plan.
Summary Table: NewsBlur at a Glance
| Detail | Information |
|---|---|
| Founder | Samuel Clay |
| Founded | 2009 (side project); full-time from 2012 |
| Headquarters | New York City and San Francisco |
| Y Combinator | Summer 2012 batch |
| Open Source | MIT License, available on GitHub |
| Post-Google Reader Growth | 1,500 to 60,000+ users (March 2013) |
| Platforms | Web, iOS (native), Android (native) |
| Free Tier | Up to 64 sites |
| Premium Plan | $36/year (1,024 sites, search, River of News) |
| Premium Archive | $99/year (full archive, Ask AI, daily briefing, 4,096 sites) |
| Premium Pro | $29/month (5-min fetching, regex, 10,000 sites) |
FAQs
What is NewsBlur and what makes it different from other RSS readers?
NewsBlur is an open-source personal news reader that allows users to subscribe to RSS feeds, newsletters, and YouTube channels, with features like full-article extraction, story change tracking, and Intelligence Training. What distinguishes it is the user-training system: readers can highlight or hide stories by author, tag, title, or full text, creating a personalized feed that reflects individual taste more than algorithmic inference. The platform also offers curated feeds built from human behavior signals reading time, saves, and trained-up stories more than click metrics.
Who founded NewsBlur and what is the company's background?
NewsBlur was founded in 2009 by Samuel Clay as a personal project while living in New York City. Clay developed the early version during commutes on the A train before transitioning to full-time development in 2012. The company received backing from Y Combinator in the summer of 2012 and is currently headquartered in New York City and San Francisco. The software is published under the MIT License and remains open source.
How did NewsBlur survive the Google Reader shutdown in 2013?
When Google announced the closure of Google Reader in March 2013, NewsBlur's subscriber base grew from approximately 1,500 users to over 60,000 almost immediately. Founder Samuel Clay documented the intense scaling challenge on the NewsBlur Blog in a post titled "Three Months to Scale NewsBlur." The rapid growth tested the infrastructure but ultimately validated the product's value to users seeking a capable RSS alternative.
What are the four curated feeds NewsBlur added in 2026?
In 2026, NewsBlur added four curated rivers to its sidebar: Global Shared Stories (stories users chose to share with commentary), Widely Read Stories (stories that held the most reading time), Long Reads (features and essays that readers spent significant time on), and Good Reads (stories that at least two people finished reading closely and then saved, shared, or trained up, with a deliberate tilt toward smaller feeds). None of the four ranks by clicks.
What AI features does NewsBlur offer?
NewsBlur's AI features are available on Premium Archive and Pro tiers. Ask AI allows users to ask questions about any story and receive AI-powered answers. Daily Briefing delivers a customizable summary of a user's top stories each day. For the Global Shared Stories feed, Claude Haiku is used to rank stories once per hour, weighing the quality of user comments attached to shares. These tools augment the reader's own curation without replacing the Intelligence Training system that remains the platform's core differentiator.