Scorchsoft
  • Free tool
  • Open source · LGPL
  • Windows
  • Linux

Quick Whisper

Dictate polished text straight into your work

Press a shortcut, speak, and Quick Whisper transcribes and optionally copy-edits your words before pasting them into the active app. Free, open-source software; bring your own OpenAI API key.

Latest version
v2.4.1Latest version
Builds
Windows & LinuxBuilds
OpenAI API
Bring your own keyOpenAI API
WHAT IT DOES

Speak naturally. Get text ready to use.

Version 2.4.1 adds push-to-talk, searchable history and a proper light theme, while making recording and paste more dependable.

Dictate with one shortcut

Use toggle recording or hold the shortcut for push-to-talk. Choose a raw transcript or AI-edited text.

Paste into the active app

Copy or paste the finished text into an email, document or message without switching away from your work.

Find and reuse earlier text

Search saved dictations, preview full entries and bring a past transcript or edit back into the main window.

GETTING STARTED

Get dictating in a few steps

The app is free. Your own OpenAI API account covers transcription and any AI editing calls.

  1. Download and open Quick Whisper

    Choose the appropriate build in Downloads and launch it. First run explains the API key and pauses hotkeys while you enter it.

    The main Quick Whisper app interface on Windows, showing the microphone input selector, transcription pane and the Record + Transcript and Record + AI Edit buttons
  2. Choose microphone and models

    Select the input device and adjust the transcription model, copy-edit model and prompt in Settings.

    Quick Whisper's Adjust AI Models window, showing the OpenAI transcription model set to whisper-1 and the copyediting model set to gpt-4o
  3. Record and process

    Press Ctrl+Alt+J for transcription with AI editing, or Ctrl+Alt+Shift+J for the raw transcript. Stop with the same shortcut in toggle mode; push-to-talk is optional in v2.4.1.

  4. Copy, paste or revisit

    Enable auto-paste to insert text into your active application, or copy it manually. Use File > Browse History to find a prior entry.

A LOOK INSIDE

Explore Quick Whisper

These images are from earlier builds; the v2.4.1 interface has updated settings, history and themes.

Main Quick Whisper interface (earlier build)
Model settings (earlier build)
Retry a recording (earlier build)
Prompt management (earlier build)
The main Quick Whisper app interface on Windows, showing the microphone input selector, transcription pane and the Record + Transcript and Record + AI Edit buttons
Main Quick Whisper interface (earlier build)
“I want to thank you for giving me back my voice. I have had a long road of having parts of my tongue removed for cancer over the years, with the last surgery being four and a half years ago and taking the remainder of my tongue, including my voice box. Not only have you given hope to people with laboured speech, which I have experienced in recent years, but you have also given people with no voice a voice. Kudos to you for adding text-to-speech to video conferencing tools like Microsoft Teams. I could never thank you enough for the gift you have given me.”

Niko

Text-to-Mic user · Member of the mute community

RELEASES

What is new in Quick Whisper

  1. v2.4.1Latest

    Push-to-talk and searchable history

    Added hold-to-record push-to-talk, searchable dictation history, a proper light theme, clearer recording and processing status, improved clipboard and auto-paste handling, grouped settings, better first-run guidance and recoverable API errors.

  2. v2.3.0

    A more reliable dictation loop

    Added a live microphone meter, elapsed timer, refreshable input list, recording limits, persistent history, retry and re-edit controls, improved hotkey feedback and safer settings storage.

  3. v2.2.1

    Updated AI models

    Updated default transcription and copy-editing model choices, added alternative models, and aligned release build names with app versions. Existing users keep their saved model choices.

Full history (v2.1.3 – v2.2.0)
  1. v2.2.0

    Memory management

    Robustness improvements to memory management and additional diagnostics for reported memory growth.

  2. v2.1.5

    Tray and window behaviour

    Fixed tray restoration and cleanup, and added an optional close-to-tray setting.

  3. v2.1.4

    Paste settings and languages

    Added paste-method choices and Japanese, Korean, Russian and Portuguese language support. This release was marked as a pre-release.

  4. v2.1.3

    Keyboard shortcut reliability

    Expired stale key presses and cleared stray key state to improve shortcut recognition; included a console build for diagnostics.

Download Quick Whisper

Windows v2.4.1 is the newest linked build. Linux has an older linked release; All versions shows every available file.

Windows

v2.4.1 · latest

EXE · 34 MBDownload

SHA-256 f366e4d6bcdabd8d…

Linux

v2.1.2 · latest

ZIP · 73 MBDownload

SHA-256 29d68d2530d4c4d4…

Build it yourself

Source code

Review and compile the code yourself — handy if your virus scanner flags the build.

View source ↗

Unsigned Windows builds may trigger a security prompt. Review the source or build it yourself if you prefer. A Mac binary is not currently linked.

24 more files: earlier versions and alternative builds.

All versions and builds

  1. v2.4.1Latest

    Added hold-to-record push-to-talk, searchable dictation history, a proper light theme, clearer recording and processing status, improved clipboard and auto-paste handling, grouped settings, better first-run guidance and recoverable API errors.

    • Windows · EXE · 34 MBSHA-256 f366e4d6bcdabd8dffa19da09420a0c541b6c37097238afafac59dc8d304599aDownload
  2. v2.3.0

    Added a live microphone meter, elapsed timer, refreshable input list, recording limits, persistent history, retry and re-edit controls, improved hotkey feedback and safer settings storage.

    • Windows · EXE · 34 MBSHA-256 827b793659bacf9d58792fadfec5e69966e86aff0f6769ec187aba4bcb638195Download
  3. v2.2.1

    Updated default transcription and copy-editing model choices, added alternative models, and aligned release build names with app versions. Existing users keep their saved model choices.

    • Windows · EXE · 34 MBSHA-256 836592a123714c24d6000b8bdcc768acb59a245c18014bbb42b3a917b3942fd6Download
  4. v2.2.0

    Robustness improvements to memory management and additional diagnostics for reported memory growth.

    • Windows · EXE · 34 MBSHA-256 4cafce98dca80c297e79e267d0a9cb6119ddc63ea9655e2688eb096baeecae13Download
  5. v2.1.5

    Fixed tray restoration and cleanup, and added an optional close-to-tray setting.

    • Windows · EXE · 34 MBSHA-256 df346947ce21b0e89edd362df7b1639fd029c3e8ebe85b1f273182f8e931ad26Download
  6. v2.1.3

    Expired stale key presses and cleared stray key state to improve shortcut recognition; included a console build for diagnostics.

    • Windows · EXE · 33 MBSHA-256 85e51381a1af70fa866bc0069c7f17d1643ff0b53dd61bdb0d5092bf30044936Download
  7. v2.1.2

    • Windows · EXE · 33 MBSHA-256 c94ef3d63f0799d4bf28806db3da921138991a2484e3d60d8d76be5c60f7fa51Download
    • Linux · ZIP · 73 MBSHA-256 29d68d2530d4c4d4f9173e6669303c68d49049966b62def07f0512bbdf3903bbDownload
  8. v1.9.3

    • Windows · EXE · 27 MBSHA-256 554e79326fa3b9f901332698be15662f5a65e35114e9aac94a16ad8797b77889Download
    • Windows · ZIP · 27 MBDownload
  9. v1.9.1

    • Windows · EXE · 74 MBSHA-256 8201c0001605daafa5359e14dca0b260de01669b54684c91fdbe427d5fc306e7Download
    • Windows · ZIP · 73 MBDownload
  10. v1.8.0

    • Windows · EXE · 72 MBSHA-256 81587a435b21e5d7080e27e04782633ab674f41664282cfa3bdb89fc6320d62fDownload
    • Windows · ZIP · 71 MBDownload
  11. v1.7.0

    • Windows · EXE · 71 MBSHA-256 4d0824432fb632110bab328bd5f4710247abbbf2bbc722f9711456cfd2eb79c7Download
    • Windows · ZIP · 71 MBDownload
  12. v1.6.1

    • Windows · EXE · 71 MBSHA-256 c31e8149356c773fba950fcc2de4352fa656380ad7f7d10bb26ac612c47cdc6dDownload
    • Windows · ZIP · 71 MBDownload
  13. v1.5.0

    • Windows · EXE · 71 MBSHA-256 5763289500ee12c939da9a7261e5e9873ada969658e84f94f930270b197a229fDownload
    • Windows · ZIP · 71 MBDownload
  14. v1.4.0

    • Windows · EXE · 71 MBSHA-256 088b67302bde0958dc6f122f9d414badbc852c8022ac7256a84a81c6780642f7Download
    • Windows · ZIP · 70 MBDownload
  15. v1.3.0

    • Windows · EXE · 70 MBSHA-256 9db5def8e2b8eb7b5b2d06f4f354d3728125248dc10e4066da49537411bb8d54Download
    • Windows · ZIP · 70 MBDownload
  16. v1.2.0

    • Windows · EXE · 70 MBSHA-256 8bec7c330ba461ee40b67c86691e75c4edd5998bf2ebcdaae5f4577b64b78e36Download
    • Windows · ZIP · 70 MBDownload

Frequently asked questions

The app is free and open source. You supply an OpenAI API key and pay for the transcription and AI editing calls it makes on your behalf.

The release adds push-to-talk, a searchable history browser, a proper light theme, clearer recording feedback and improvements to clipboard, auto-paste, settings and error handling.

Yes. Use Ctrl+Alt+Shift+J for transcription without the extra copy-editing step. Ctrl+Alt+J transcribes and edits.

The Downloads section currently includes a Linux v2.1.2 build and Windows v2.4.1. There is no linked Mac binary; the GitHub repository has source and build instructions.

The app saves configuration locally on your computer. Treat the API key as a secret and do not share it; OpenAI charges your account for its use.

Want to talk about your project?

Tell us what you’re trying to achieve and we’ll map the fastest credible path.