Meet Leo — your AI reading assistant

Stay informed.
Without the noise.

Newsfeed subscribes to the RSS sources you care about, then Leo — our AI reading assistant — filters out the noise and summarizes what's left. Power readers save 6+ hours a week.

6,200+
power readers
84K+
feeds subscribed
12M+
articles filtered
6.3 hrs
saved weekly, avg.
📰AI Research · Today
Leo on
9
Import AI · 12 min ago
Anthropic releases Claude 4.6 with 1M context — benchmarks
↳ matched "AI research"
8
The Information · 1 hr ago
OpenAI's next model delayed to Q3, insiders say
↳ matched "AI industry"
8
Stratechery · 3 hrs ago
Aggregation theory, revisited in the age of AI agents
↳ matched "AI strategy"
4
Hacker News · 4 hrs ago
Show HN: I built a weekend project with three AI agents
⌵ below threshold
2
TechCrunch · 5 hrs ago
Celebrity gossip disguised as tech news, again
⌵ below threshold
3 of 47 articles surfaced by Leo todaysaved 38 min
What Leo does

Six features. Zero fluff.

We stripped the RSS reader down to what actually helps you read less and know more. Everything else got cut.

Leo ranks every article 0-10

Your topics in, relevance score out. Articles below your threshold collapse into a 'More' drawer — they don't disappear, they just stop screaming.

Three-sentence AI summaries

Click any headline for a Claude-generated TL;DR. Decide in 10 seconds whether to read the full piece.

Topic boards, not folders

Group feeds into boards like 'Competitors', 'AI Research', or 'Local politics'. Each board keeps its own relevance threshold.

Distraction-free reader

Cheerio strips ads, popups, and related-post clutter. Just the article. Save for later with one tap.

OPML import in 30 seconds

Coming from Feedly, Inoreader, or NetNewsWire? Drop your OPML file and every feed lands pre-organized.

Morning Brief email

The top 8 articles across all boards, summarized, in your inbox at 7 AM. Skip the app entirely if you want.

How it works

From RSS firehose to focused feed in 60 seconds.

01

Tell Leo your topics

Pick from AI, SaaS, Fintech, Climate, Marketing — or type your own. Leo uses these to score every article.

02

Add feeds or import OPML

Paste any RSS URL. We auto-fetch the favicon, title, and last 20 articles. Or drop an OPML file.

03

Read only what matters

Feed sorted by relevance, not chronology. Summaries on demand. Save, skim, or skip.

Time saved

Power readers reclaim their mornings.

Three stories from people who used to live in Feedly and now spend their news budget on Leo instead.

5.5 hrs/day saved
I went from 4 hours of RSS triage a day to 40 minutes. Leo caught 3 competitor launches I would have missed in the firehose.
Priya Natarajan
Priya Natarajan
Senior Analyst, Redpoint Ventures
30 hrs/mo saved
Our product team has a 'Competitors' board with 42 RSS feeds. Newsfeed surfaces ~8 relevant articles a day. Before, we drowned.
Marcus Whitfield
Marcus Whitfield
Head of Product, Linear
14 hrs/wk saved
As a solo founder I can't afford to miss industry shifts, but I also can't spend my morning in Feedly. The 7 AM brief is my entire news diet now.
Hana Oyelaran
Hana Oyelaran
Founder, Stitchwise
Pricing

No free tier. Real AI costs money.

Every plan includes Leo. 14-day refund, no questions. Change or cancel anytime.

Reader

For individuals who read the news every morning.

$7/ month
billed annually
  • Up to 100 feeds
  • 3 topic boards
  • Leo AI relevance scoring
  • 3-sentence summaries (unlimited)
  • Morning Brief email
Start as a Reader
★ Most popular

Pro

For analysts, researchers, and founders.

$15/ month
billed annually
  • Unlimited feeds
  • Unlimited boards
  • Full-text article extraction
  • OPML import + export
  • Trending topic alerts
  • Priority Leo processing
Go Pro

Team

Per seat. For product and research teams.

$32/ month / seat
billed annually
  • Everything in Pro
  • Shared team boards
  • Slack + Teams digests
  • Admin + billing controls
  • SSO (SAML)
Start Team trial
FAQ

Frequently asked

Yes. Every article Newsfeed ingests is passed to Claude Haiku, which scores its relevance to your declared topics on a 0-10 scale. Summaries are generated on demand to keep latency and costs reasonable.