Prose style. Tone. Perspective.
Your voice.
A multi-dimensional voice system — prose style, tone, POV, and dialogue balance — that makes your writing yours.
The Slow Ache
Craft patterns, vocabulary, and formatting preferences that shape every chapter draft.
- — Favor em-dashes over commas for interruptions and pivot points
- — Repeat key sensory details across scenes as motifs
- — Paragraphs end mid-beat — deny resolution at the sentence level
- — No exclamation marks; intensity comes from rhythm, not punctuation
Unhurried, sensation-forward, steeped in longing — over the character's shoulder. Earn the intensity through specificity and restraint — not through purple prose or cliche romance language.
She traced the rim of her glass with one finger, the cool condensation slick under her skin, while he sat across the scarred oak table, shirt sleeves rolled to his elbows. A vein pulsed at his wrist, faint blue against the tan. Her chest pulled tight, breath shallow in the quiet kitchen, where the fridge hummed low and the window framed dusk bleeding orange into the yard. He shifted, boot scraping the linoleum, and his knee nudged the table leg — close enough that she felt the vibration up her own chair. Not her. Not quite. His hand lay open on the wood, palm upturned, a smudge of grease still darkening one knuckle from the garage earlier.
A voice sample — not story content.
AI fiction has a voice problem.
Voices change this.
A multi-dimensional voice system that defines how your story is told — not as a suggestion, but enforced in every chapter the AI drafts.
Prose style. Tone. POV. Dialogue. One voice.
Every voice is built from layered dimensions — eight prose styles from confessional to cinematic, seven tones from tender to dangerous, three narrative perspectives, and three dialogue balances. Each combination produces a distinct narrative texture that shapes how every chapter sounds. Start from nine curated presets or build your own from scratch.
Every silence means something. Every almost-touch is a sentence.
Two hands on the same counter, almost touching. Close enough to feel the warmth. Neither of them moved. The coffee was getting cold. Neither of them cared.
Paste your prose. Meet your voice.
Already have a writing style? Paste 200 or more characters of your own prose and the style analyzer classifies it across all four dimensions, explains its reasoning for each choice, generates a fingerprint of craft patterns the presets don't cover, and suggests a voice name. Review everything, adjust what you want, and save a voice that sounds like you.
I shouldn't have opened the door. I knew that. But it was already open and the hand was on the frame and I was already backing up, and the thing about want is that it doesn't ask permission — it just moves your feet before your brain catches up...
Stream-of-consciousness rhythm, self-correcting sentences, real-time processing
Blunt emotional honesty, no ornament, feelings stated directly
Direct "I" narration, deep interiority, narrator as participant
No dialogue present; interiority and action carry the passage
- — Run-on sentences as breathlessness — commas where periods would be, subordinate clauses stacking
- — Self-contradicting rhythm: assertion, immediate undercut, then doubling down
- — Metaphor grounded in physical sensation — want is rendered as movement, not abstraction
- — No semicolons, no em-dashes for parenthetical asides — commas only, mimicking spoken cadence
The patterns presets can't capture.
Presets handle prose style, tone, perspective, and dialogue balance. But your voice has texture the categories miss — sentence mechanics, vocabulary register, rhetorical habits, punctuation style, the things you deliberately avoid. The style fingerprint captures those patterns as actionable craft instructions, injected as high-priority voice characteristics into every chapter draft. Generated by the analyzer or written by hand.
These set the shape — prose texture, emotional register, narrative distance, and scene rhythm.
- —
Favor em-dashes over commas for interruptions and pivot points
Sentence mechanics - —
Repeat key sensory details across scenes as motifs (a specific scent, a gesture, a phrase)
Rhetorical habits - —
Paragraphs end mid-beat — deny resolution at the sentence level
Sentence mechanics - —
No exclamation marks; intensity comes from rhythm, not punctuation
Anti-patterns - —
Prefer concrete body language over named emotions — show the flinch, not 'she felt afraid'
Vocabulary register
The details that make it work
Frequently Asked Questions
A voice is a multi-dimensional profile that defines how your story is told. It combines a prose style (like Confessional, Cinematic, or Deep Slow Burn), a tone (like Aching, Raw, or Playful), a narrative perspective (first, second, or third person), and a dialogue balance. Every chapter the AI drafts is shaped by these dimensions together — not as a suggestion, but as an enforced constraint.
There are eight prose styles — from literary to pulp, confessional to cinematic — and seven tones — from tender to dangerous. Each combination produces a distinct narrative texture. "Raw" + "literary" reads entirely differently from "raw" + "pulp." The system resolves cross-dimension interactions deliberately, so the output genuinely changes with every combination.
The style analyzer lets you paste 200 or more characters of your own writing and have the system classify it across all four voice dimensions. It explains its reasoning for each choice, generates a style fingerprint of craft patterns the presets don't cover, and suggests a voice name. Review everything, adjust what you want, and save a voice that sounds like you.
A style fingerprint captures the craft patterns that presets can't — sentence mechanics, vocabulary register, rhetorical habits, punctuation style, and things you deliberately avoid. These patterns are injected as high-priority voice instructions into every chapter draft. You can generate a fingerprint from the style analyzer or write one by hand.
Yes. The prose preview generates a sample passage using your current voice settings so you can hear how the combination of prose style, tone, perspective, and dialogue balance sounds together — before committing to a single chapter.