Alpha — early access. This page describes an early-stage product; features and pricing may change.
Vol. I · A drafting instrument for independent writers

Quillecho

Your archive already sounds like you.
Now your drafts can too.

The lead

Quillecho reads your published back-catalog and drafts new pieces in your own voice — so the blank page starts closer to finished, in the cadence and word choices that are actually yours.

Quillecho ingests your existing published work — a Substack or beehiiv archive, a personal blog, or a folder of freelance bylines — and builds a private voice model from it. From then on, you give it a topic or a rough outline and it produces a first draft written the way you already write: your sentence rhythm, your typical structure, your recurring phrases and tics. You edit from there. It doesn't publish for you, and it doesn't pretend to be a person — it's a drafting tool meant to get you further than a blank page — we're not publishing a specific accuracy or completion percentage because no voice-match benchmark exists to back one.

Who it's for. Independent, single-author newsletter and blog writers who publish under their own byline (not as staff under an employer's editorial policy), have a back-catalog of at least ~20 published pieces in one place, and write on a recurring weekly-or-biweekly cadence.

Think: Substack and beehiiv writers, Ghost(Pro) bloggers, freelance columnists and trade-press contributors with a consistent byline, LinkedIn native-newsletter creators, and ghostwriters who maintain their own newsletter as a portfolio piece.

Explicitly not for: staff journalists writing under an employer's editorial and AI-disclosure policy. Individual newsrooms increasingly set their own AI-disclosure rules, and a staff writer doesn't own that call the way an independent one does — so we build for the writer who owns their own byline and their own disclosure decision. This is a deliberate ethics fence, not a market-size accident, and it stays even though it narrows who we can sell to.

The specimen

A simulated walkthrough of the three moments that matter: connecting an archive, generating a voice-matched draft with a side-by-side comparison, and running a revision loop. Every scenario is clearly labeled SIMULATED — this is a mockup of the product experience, not a live tool.

Recto & verso

The demo below is set as a facing spread — your archive on one page, the drafted piece on the other. Open it full-page for the actual side-by-side comparison.

How it works

Archive ingest

Connect one source at launch — paste in a Substack/beehiiv export, an RSS feed, or a batch of URLs. Your voice model builds from the pieces you actually published.

Voice-matched drafts

Give it a topic or outline; get a full first draft back in your cadence, not a generic "helpful assistant" tone. Unlimited drafts on every tier.

Revision loop

"Make it punchier," "more like my March piece," "cut the throat-clearing intro" — a plain-language revision pass that nudges the draft closer without starting over.

Style-drift alerts

Flags when a draft is reading generic or off-voice before you spend time editing it — a heads-up, not an automated "voice-match score" (see the demo for why we don't claim one).

Multiple sources

Blend a Substack archive with freelance bylines published elsewhere into one voice model, for writers whose work lives in more than one place.

Your read is the judge

No invented accuracy score. The evaluation is you, reading the draft side-by-side with your own archive — the demo shows exactly what that comparison looks like.

The price

Starter — $24/mo
One archive connection, voice-model build, unlimited drafts, 1 revision loop per draft.
Pro — $49/mo
Multiple archive sources, priority draft turnaround, style-drift alerts.

Anchored to HyperWrite's Custom Personas pricing ($16–45/mo across annual/monthly plans) — the closest functional comparable found. No team/enterprise tier at launch; this is built for one author, one voice.

Rough payback, illustrative: at $24/mo, Quillecho pays for itself if it saves you roughly 20–30 minutes per issue, valuing your own writing time conservatively. For a weekly newsletter that's under 2 hours a month of recovered time. (unverified)

Letters & hard questions, answered honestly

Isn't this just AI slop with my name on it?

The whole point is that it isn't generic — it drafts from your own archive, not a general internet-average voice. We don't hide AI involvement either; we recommend you decide your own disclosure practice rather than pretending nothing happened.

How do you actually know it sounds like me?

Honestly: there's no benchmarked "voice-match score," and we won't invent one. The demo shows a side-by-side of your archive excerpt against the drafted piece — you're the judge. The revision loop exists because a first pass won't always land; the job is to get closer with your feedback, not claim perfection on try one.

What happens to my archive — do you train a model on me, and who owns that?

Your archive is used only to build your own draft generation — it is not pooled or shared across other users' models, and it's deletable on request. This is a plain policy commitment, not a technical claim we haven't built yet.

Will my publication or employer fire me for using this?

This is exactly why Quillecho is built for independent writers who own their own byline and disclosure decision — not staff journalists bound by an employer's AI policy. That group is explicitly out of scope for us.

Why pay $24–49/mo instead of just pasting my archive into ChatGPT or Claude myself?

You can, and some writers will. The honest edge isn't a smarter model — it's not re-pasting 20 pieces every session, a revision loop built for this exact task, and drift alerts that flag when a draft is going generic. It's a workflow wrapped around a general model, the same way HyperWrite, Jasper, and Copy.ai compete against "just use the base model."

Want in early?

We're looking for a small number of independent writers to try Quillecho against their real archive before anything is publicly available.

Request an early-access slot
contact: [placeholder — add real inbox before this page goes anywhere real]