How Letterfork Keeps Posts in Your Own Voice

Letterfork keeps your voice by training on your own past newsletters before it writes anything. You drop in 5–10 previous issues once, and the tool extracts your sentence rhythm, opening habits, vocabulary, and humor. Every LinkedIn post, X thread, Bluesky post, Substack Note, Threads post, Instagram caption, and Reddit draft it generates is then shaped to read the way you already write — not the way a generic AI template writes.

That distinction matters most if you've tried other repurposing tools and abandoned them because the output sounded like everyone else's.

Why generic AI repurposing loses your voice

Most repurposing tools start from a blank model with no memory of you. They compress your newsletter into platform-sized chunks using default phrasing, default transitions, and default hooks. The result is technically your content but not recognizably your writing — the rhythm flattens, the openings turn generic, and the humor disappears.

Letterfork's approach inverts that: the voice model is built first, from your material, and the repurposing happens inside it.

The three steps, and where voice enters

1. Train your voice (one time)

Drop in 5–10 of your past newsletters. Letterfork extracts:

• Sentence rhythm — how long your sentences run, where you break them
• Opening habits — how you tend to start a piece
• Vocabulary — the words and phrases you actually reach for
• Humor — your register and timing

This is a one-time setup. You don't retrain for every issue.

2. Paste a newsletter URL or full text

Give Letterfork the issue you want to repurpose. It works with Substack, Beehiiv, and Ghost.

3. Get platform-native output

One issue becomes posts for seven platforms: LinkedIn, X, Bluesky, Substack Notes, Threads, Instagram, and Reddit. The site claims this takes about 60 seconds.

What "in your voice" actually looks like

The site's own example shows the same newsletter excerpt reshaped into a LinkedIn post and an X thread. The LinkedIn version keeps the writer's blunt framing ("glorified data entry on my own words") and the specific detail about talking to thirty other writers. The X thread breaks the same material into numbered beats (1/4) while preserving the same vocabulary and the same dry, exasperated tone.

That's the test to apply to any output: does it still sound like the person who wrote the newsletter, or like a summary of them?

How to judge it for yourself

• Run it on an issue you're proud of. If the output preserves your distinctive lines and cadence, the voice model is working.
• Compare across platforms. The tone should stay consistent even as the format changes — a thread is structured differently from a caption, but it should still sound like you.
• Check the humor. Humor is the hardest thing to carry over and the fastest signal that a tool has reverted to generic phrasing.

Practical notes

• Free tier: The first three rewrites are free, no credit card required.
• Cancellation: The site states you can cancel any time.
• Sources supported: Substack, Beehiiv, Ghost.
• Pricing beyond the free tier: Not specified in the available material — check the pricing page for current details.

Who this fits

Letterfork is aimed at newsletter writers who already enjoy the writing but resent the "repurposing tax" — the hours spent reshaping one piece for multiple platforms. If your problem is that repurposed posts don't sound like you, the voice-training step is the specific mechanism to evaluate. If your problem is simply volume, any repurposing tool will do; the voice model is what separates this one from the generic options.