auris.
Features

The spec sheet.

Every feature in Auris, grouped to skim. Jump to a section.

Capture · 001

The dictation loop.

Hotkeys, HUD, hands-free toggle.
The plumbing that gets your voice into Auris.

Hands-free toggle.

Tap to start, tap to stop. 

Or stay quiet and auto-stop kicks in.

Replay last dictation.

Re-inject the most recent polished output with a third hotkey.

Dictation HUD.

A small indicator pulses with your mic level.
Click-through, multi-monitor aware.

Mic handling.

Any input device. Survives USB reconnects. 

Respects the OS privacy toggle.

Polish · 002

Raw transcript to clean text.

Pick a model. Move three sliders. 
Toggle the cleanups you want. Watch what happens.

Three tone sliders.

Formality, conciseness, assertiveness.
Five positions each.

Seven cleanup toggles.

Filler words, capitalization, smart quotes, em-dashes, and more.

Five output formats.

Plain prose, bullets, numbered, headed, or app-driven.

Role-break guard.

If the model tries to answer instead of polish,
Auris drops the output.

Polish bench.

Test your settings on any pasted text without dictating.

Vocabulary · 003

Words it learns.

Dictionaries, snippets, domain packs, brand voices.
Auris speaks your dialect.

Custom dictionary.

Add words manually, or
let Auris learn from your edits.

Snippets.

Spoken triggers expand to longer text.
Variables for date, time, clipboard.

Six domain packs.

Healthcare, Legal, Finance, SEO, Tech, YouTube.
Pin one as default.

Brand voices.

Upload a Markdown style guide.
Switch between voices per project.

App-aware · 004

Polish per app.

Different settings switch in based on which app has focus.
Five modes, four dropdowns, plus free-text overrides.

Four dropdowns per app.

Mode, tone, format, language. Stack them per app — defaults cover most, override the rest.

Per-app instructions.

Free-text override field. "Always sign with my name." "Never use exclamation points." Anything the dropdowns don't cover.

Auto-switch.

Auris reads which app has focus and applies its preset automatically. No manual selection.

Compose · 005

When dictation is the destination.

For when you're not pasting into another app. Memos as real files. Studio Mode for long-form.

Memos.

Long-form notes saved as Markdown files in a folder you control.

Real files on disk.

No export step because there's nothing to export from.

Recall · 006

Past dictations, measured.

Every dictation is searchable. Every minute saved is countable.
All on your machine.

History.

Every dictation, raw and polished, with app, timestamp, and provider.

Inline editing.

Edit polished text in place. Re-polish with new settings.

Range filters.

Slice by date, app, or text search.

Day streak.

Consecutive days dictated, with personal best.

Privacy · 007

Flip per session.

Redact patterns, force offline, audit the network, archive or delete.

Pattern redaction.

Email, phone, SSN, credit card. Add your own regex.

Data flow board.

Live view of what travels where with your current settings.

Network audit log.

Every outbound call, with destination and bytes.

Export everything.

One-click archive of your full data set.

Wipe everything.

One-click destructive reset. No undo.

Customize · 008

Themes, setup, reliability.

Twelve themes, a re-runnable wizard, tray-resident persistence.

First-run wizard.

Provider, API key, typing speed, theme.
Re-runnable any time.

Hotkey live-reload.

Change the binding; it takes effect immediately.

Test connection.

Verify your LLM and STT providers with a round-trip latency check.

Whisper model picker.

Tiny, Small, Medium, Large as on-demand downloads.

Auto-start on login.

Optional. Sets the tray icon, doesn't open the dashboard.

Tray-resident.

Closes to the tray. The dictation pipeline keeps running.

Ready to run it?

Auris is currently in beta until early June.
30-day money-back, no questions.

Still in Beta How it works