Ledenote builds the Omnibox — answers from your own archive, in every language your readers type, embedded in your brand.
Five capabilities, each mapped to a metric on your dashboard — inventory yield, click retention, editorial line, brand integrity, time-to-live.
Native ad slots live inside the conversation flow. Every impression is tracked, every click attributed — same metrics your ops team already exports, minus the friction of an integration.
AI summaries already answer your readers. They cite your reporting and pocket the click. Ledenote answers them on your masthead — every paragraph cited to your articles, every follow-up routed deeper into your back catalog. The next click stays with you.
Hard refusals on topics your masthead doesn't address — competitor coverage, legal-risk territory, anything off your editorial line. The Omnibox declines politely. It doesn't invent a workaround.
One <script> tag, on a staging URL the same day. Your CTO reads the SOC 2 doc; your editor configures everything else.
Your logo, your colors, your typography, your tone of voice. The Omnibox arrives styled as part of your site — readers don't recognize it as a third-party tool, and the request never leaves your domain.
A reader looking for a name isn't asking the same thing as a reader catching up on a week-long story. The Omnibox reads the situation and shapes the answer to match.
A name, a place, a topic. The Omnibox returns a 3–4 bullet sketch and the matching articles you've filed on it — fast enough to replace your site search.
Factual questions ('who won?', 'when does Ramadan start?') get a grounded paragraph, citations to your coverage, and 1–2 follow-up questions to keep the reader inside your archive — nothing invented, nothing extra.
Complex stories don't fit one paragraph. The Omnibox stitches a quick recap, a chronological timeline, and 2–3 follow-ups — every link going back to your coverage.
Each Omnibox is built around your editorial language. It meets every reader where they type — and delivers every answer where your editors wrote.
We're picking five publishers as design partners through 2026. We wire up the integration, you ship in 24 hours, and your input shapes the roadmap. Your founder rate stays locked for the lifetime of the contract.
The questions an editor-in-chief asks before a 30-minute call — answered straight, at the capability level.
The Omnibox replaces a news site's native search with cited, multilingual answers drawn only from that newsroom's own archive — on the publisher's brand, live in 24 hours.
We point a staging Omnibox at your sitemap before the first call, and a production widget is live in about 24 hours.
Readers can ask in any language; the Omnibox answers in the newsroom's publishing language, citing the original articles.
Yes — the widget is paywall-aware and respects reader authentication and entitlements.
Every answer is built only from the newsroom's own archive, with each paragraph cited back to the source articles.
A drop-in <script> tag, with a programmatic API and SSE streaming available for deeper integrations.
Three tiers — Newsroom, Masthead, and Media Group — with founding-cohort rates. See the pricing page for current numbers.
Each capability below ships as a documented REST endpoint — full OpenAPI spec live inside your dashboard from day one.
RSS, signed webhook push, or scheduled sitemap crawl — pick whichever your CMS already speaks. New posts and corrections propagate the same way. No batch jobs, no nightly catch-up windows.
Token-level streaming over standard HTTP. Sub-400ms time-to-first-token. No websockets, no proxy surprises, works behind every CDN your readers already hit.
Pass your reader's identity through with a signed JWT. Subscriber-only answers, anonymous fallbacks, hard cutoffs at quota — wired into your existing sign-in flow, not replacing it.
Each Omnibox indexes only the articles you publish — no shared corpora, no pooling with other newsrooms. Full export on request; deletions propagate before the next query.