Free gets you the ranked front page. Premium staffs it — an AI desk that recaps, briefs, answers and traces, grounded in today's stories and nothing else.
Every story gets an original summary remade from the source — with the ranking receipt attached.
A newly released open-weights model has overtaken closed rivals on three major coding benchmarks, reshaping the build-vs-buy maths for AI teams.
Your top five stories, synthesised into one read — what happened, why it matters, what to watch.
Today turns on three things: the ceasefire talks entering their decisive day, an open-weights release scrambling the AI race, and a chip earnings beat that says the demand story isn't done. Watch the talks — everything else prices off them…
A conversational desk grounded in the live feed. Every answer cites the stories it came from.
Trace today's headline back through the beats that lit it — oldest first, ending at NOW.
First open-weights frontier release proves the approach viable.
Three labs commit to open releases; benchmark gap narrows to single digits.
Enterprise adoption of open models passes 40% in new deployments.
An open-weights model leads all three coding benchmarks for the first time.
No — and that's deliberate. The desk works from headlines, publisher descriptions and your feed context, then writes original recaps. Nothing is scraped or reproduced; the "Read full story" link always goes to the publisher.
The paper still prints. Every AI surface falls back gracefully to publisher snippets — you never hit a broken page, just a quieter one.
No. Your signals stay on your account and only tune your own ranking. The "why you're seeing this" receipt shows exactly what they did.
Any time, in one tap from Profile. You drop back to Free and keep your tuned feed — the ranking is yours, not Premium's.