Speak. AI listens.
Your prompt is ready.

Dictate and get clean text for any apps, a polished rewrite, or a ready-made prompt for ChatGPT or Claude — inserted right into the active field. Transcription works offline.

Used by 1,008+ people · Windows · macOS · Linux

Rubic app dashboard

Sound familiar?

Dictated — then spent an hour editing

Voice notes can't just be pasted into ChatGPT

Writing prompts manually every time

Same thing, over and over, 5 minutes per prompt

Switching between windows

Copy, switch, paste — lost my train of thought

01 / TRANSCRIPTION

Speak — get text instantly

Press the hotkey, dictate, release. Text appears in the active field — no copying, no window switching. Works offline with a local model or online for maximum speed.

Local model — works offline, data stays on your device
Online mode — 3-4x faster (available in Pro)
Insert into active input — right where you work
100+ languages supported
New Message
ToSsarah@company.com
SubjectFollow-up: Q4 proposal review
⌥ Space to dictate

Hold ⌥ Space → speak → text inserted into the active field

Removed filler
Added punctuation
Fixed structure

Raw input

so basically we've been looking at the onboarding flow uh and I think the main issue is users are dropping off at step three because there's too much information all at once um we should split it into smaller steps and add progress indicators so people know where they are uh yeah that would help a lot

AI processed

Clean output

Basically, we've been looking at the onboarding flow and I think the main issue is users are dropping off at step three because there's too much information all at once. We should split it into smaller steps and add progress indicators so people know where they are. That would help a lot.

BIS

02 / POST-PROCESSING

Pro

Raw dictation → clean text

Automatically removes filler words, adds punctuation, formats paragraphs. You speak naturally — get a polished document.

Removes 'um', 'uh', and pauses
Auto headings and paragraphs
Preserves meaning without losing style
Customizable processing rules
Try Live Demo

03 / PROMPT MODE

Pro

Voice → ready prompt for any AI

Not just transcription. Prompt Mode understands your request and formats it as a professional prompt — with role, task, and output format. Just say what you need.

claude — whisp prompt mode
⌥ SpaceHold to speak — listening…

Hold ⌥ Space → speak → structured prompt inserted into any app

How it works

Three seconds from voice to result

01

Press hotkey

⌥ Space to start recording or Control + Space to prompt mode

02

Speak

Transcription runs in the background

03

Done

Text inserted into the active field

Simple pricing

Start free. Upgrade when ready.

Free

  • Unlimited transcription (local model)
  • Instant dictation
  • Limited personalization
  • Insert into active input
  • Works anywhere (Slack, Gmail, Cursor, etc.)
  • 100+ languages supported
Download Free
Popular

Pro

$10/mo

Everything in Free, plus:

  • Online transcription (unlimited)
  • Prompt Mode — voice to ready prompt
  • Post-processing — clean text
  • Premium local model (2x accuracy)
  • Full personalization across apps and tasks
  • Optimized speed and reliability
  • Increased dictation length
  • Priority support

No hidden fees · Cancel anytime · Data stays on your device

Frequently asked questions

What people say

Join thousands of people who dictate smarter every day

I replaced three different apps with Rubic. Transcription, cleanup, and prompt creation — all in one shortcut.

MR

Michael R.

Freelance Writer

The offline mode is a game changer. I work from cafes with spotty WiFi and Rubic never skips a beat.

SL

Sarah L.

UX Designer

450 minutes of online transcription per month is more than enough. I use maybe half and the quality is incredible.

JT

James T.

Podcast Producer

Start today

Try it right now

Free. Works offline.

Windows · macOS · Linux · v1.0.4