⚡ Hotkeys

AI text refiner. Voice-to-text. Audio + screen + GIF + macro recorder. Whiteboard. File transcriber with speaker labels. All from one system-tray icon.

Select text anywhere → press a hotkey → AI rewrites it and pastes it back in under a second. Speak into any text field. Record your screen. Sketch on an offline whiteboard. Transcribe podcasts with speaker labels. Edit audio. All from one tiny tray icon, all offline-first, all keyboard-driven.

Legacy builds: v3.0 Windows · v2.0 Windows · v1.0 Windows

---

What is this?

Most AI writing tools make you open a browser tab, paste your text, wait, copy the result, switch back, and paste again. That's 7 steps, every single time.

Hotkeys does it in 1 step, without leaving whatever you're writing in. And it's not just text refinement: it's a full keyboard-driven productivity toolkit that sits quietly in your system tray.

✍️ AI Text Refiner	Select any text, press a hotkey, it's rewritten and pasted back in under a second
📚 Prompt Library	16 built-in prompts + unlimited custom prompts, each with its own hotkey
🎙️ Voice to Text	Hold a hotkey, speak, your words appear wherever your cursor is. Fully offline
📝 Quick Notes	Resizable floating scratchpad with image-paste OCR and auto-save
🎨 Whiteboard	Offline Excalidraw with infinite canvas, hand-drawn aesthetic, all keyboard-driven
🎬 Screen Recorder	Capture any window or region as MP4 with optional mic audio
🎞️ GIF Recorder	Record any region as an animated GIF with adjustable FPS and duration
🔴 Macro Recorder	Record any sequence of clicks and keystrokes; replay with one key
📸 AI Screenshot	Capture, ask the AI what's in it, get an answer
🎵 Audio Editor	Built-in waveform editor for trimming, mixing, exporting audio
🎤 File Transcriber	Drop an MP3 / MP4 / YouTube link; get a transcript with optional speaker labels

No API key needed. Cerebras + Groq keys are baked in. Just download and run.

---

The Prompt Library

All your prompts in one place. Click any card to activate it, drag to reorder, right-click for options:

Default prompts (assignable to any hotkey)

Hotkey	Prompt	What it does
F1	Refine	Fixes grammar, spelling, and clarity, same meaning, natural tone
F2	Improve & Expand	Makes your text more articulate, detailed, and expressive
F3	Translate	Translates in place (default: Arabic, change to anything)
F4	System Prompt	Reformats text into a clean, deployable AI system prompt
F5	Simplify	Strips jargon and complexity, immediately understandable
F6	Technical Depth	Adds precision and implementation detail for expert readers
F7	Expand	Develops underdeveloped ideas without going off-topic
F8	Professional	Rewrites in polished, formal language at the same length
F9	Ask Claude	Turns vague thoughts into a sharp, specific AI prompt
F10	Pirate	Rewrites with nautical flair, same meaning, more swagger
—	Brutally Honest	Says exactly what's meant, no softening, no padding
—	Story Hook	Turns any idea into a gripping opening line
—	Devil's Advocate	Argues the exact opposite with equal conviction
—	Haiku	Distils the core idea into a 5-7-5 haiku
—	ELI5	Explains anything as if the reader is five years old
—	Tweet	Compresses the sharpest idea into one punchy tweet

These are the starting point. Hit + Add to write your own.

The only limit is your imagination. Want a prompt that rewrites emails in your exact voice? Converts meeting notes into action items? Translates to your language? Summarises legal contracts in plain English? Turns rough ideas into job postings?

Add it in 10 seconds and it works everywhere, instantly.

Tip: Press any per-prompt hotkey to see a floating sticky note preview. Edit the prompt on the fly before it fires, then press the key again to apply.

---

How to install

⊞ Windows, one click, no setup

1. Download Hotkeys-v3.1-win64.zip
2. Extract the zip anywhere outside Downloads/ (e.g. C:\Hotkeys\)
3. Double-click Hotkeys.exe
4. The ⚡ icon appears in your taskbar tray, you're done

No Python. No pip. No API key. No installing anything.

If your antivirus pops up (AVG / Defender / etc.): the build is unsigned for now (signed builds via SignPath are pending approval). Right-click → restore + add exception. One-time, then it's clean.

🍎 Mac, plug and play

1. Download Hotkeys-mac.dmg
2. Open the DMG, drag Hotkeys.app to your Applications folder
3. Double-click Open Hotkeys.command inside the DMG (bypasses macOS quarantine)
4. Grant Accessibility permission when prompted (one-time, 30 seconds)
5. The ⚡ icon appears in your menu bar

Prefer to install from source? Download install_mac.command, right-click → Open, terminal does everything automatically (~10 min).

---

How to use it

Refine any text (AI rewrite)

1. Select text in any app
2. Press Alt + Shift + W
3. Wait ~0.5 seconds, text is rewritten in place

Use a prompt from the library

1. Select text in any app
2. Press the prompt's hotkey (e.g. F5 for Simplify)
3. A sticky note appears, read the prompt, edit it if you want
4. Press the same key again to fire it

Add your own prompt

1. Press Alt + Shift + E to open the Prompt Library
2. Click + Add
3. Give it a name and write your instruction
4. (Optional) Assign a hotkey via right-click → Assign hotkey
5. Click Save, available everywhere immediately

Dictate text (voice to text)

1. Place your cursor where you want the text
2. Press Ctrl + Enter to start recording
3. Speak naturally
4. Press Ctrl + Enter again to stop, your words appear instantly

Runs fully offline. Whisper model bundled.

Quick Notes (floating scratchpad)

1. Press Shift + F7, a resizable notes window opens
2. Type your note, paste anything from the clipboard (incl. images → OCR to text)
3. Drag any edge to resize, drag the title bar to move
4. Press Shift + F7 again (or Esc) to close, notes auto-saved

Whiteboard (offline)

1. Press Shift + F8, an Excalidraw whiteboard opens
2. Sketch with mouse / pen / touch, drag-and-drop images, draw arrows and shapes
3. Works fully offline, no internet, no account, no telemetry
4. Press Shift + F8 again to close, scenes auto-saved

Transcribe a file (with optional speaker labels)

1. Press Shift + F9, the Transcribe tab opens in the Library
2. Drop an MP3 / MP4 / WAV / M4A / MKV / WebM, or paste a YouTube URL
3. Pick a Whisper model (base, small, large-v3-turbo, large-v3) and language
4. Toggle Speaker labels to add diarization (Speaker 1, Speaker 2, ...)
5. Click Transcribe, export as TXT, SRT, VTT, PDF, or DOCX

Audio editor

1. Press Shift + F10, the bundled audio editor opens
2. Drop an audio file onto the waveform to load it
3. Trim, splice, fade, mix, export

Record a macro

1. Press Shift + F1 to start recording, a red pill appears
2. Do anything: type, click, switch windows, scroll, every step is captured
3. Press Shift + F1 again to stop
4. Press Shift + F1 once more to replay
5. Press Esc to cancel or stop playback
6. (Optional) Name + save + assign a hotkey to frequently-used macros

Record your screen

1. Press Shift + F2, choose Full screen, a window, or drag to select a region
2. (Optional) Enable mic audio
3. Click Start Recording, dialog disappears, recording begins
4. Press Shift + F2 again to stop, save dialog appears
5. Save anywhere, MP4 by default

Record a GIF

1. Press Shift + F3, choose a window or region
2. Set FPS and max duration
3. Press Shift + F3 again to stop
4. Preview, save, or discard

AI Screenshot

1. Press PrtSc, the screen is captured
2. A dialog appears with the screenshot and an AI chat box
3. Ask anything: "What does this error mean?", "Summarise this page", "What's in this chart?"

---

Settings

Everything is configurable. Open the library (Alt+Shift+E) and click the gear icon, or right-click the tray icon.

• AI Provider, switch between Cerebras and Groq, or enter your own API key
• Hotkeys, change any global shortcut to whatever you prefer
• Voice model, choose Whisper base (fast), small (default), large-v3 (best), pick your microphone
• Transcription, language, beam size, custom vocabulary, diarization on/off
• Push-to-talk, hold to record, release to transcribe
• Autostart, launch automatically when your computer starts

---

All default hotkeys

Action	Shortcut
Refine selected text with AI	Alt + Shift + W
Open Prompt Library	Alt + Shift + E
Undo last refine	Alt + Shift + Z
Per-prompt hotkeys	F1 to F10 (assignable)
Start / stop voice recording	Ctrl + Enter
Cancel / stop anything	Escape
Macro recorder	Shift + F1
Screen recorder	Shift + F2
GIF recorder	Shift + F3
Ask Claude about selection / image	Shift + F4
Web search prompt	Shift + F5
Run a prompt chain	Shift + F6
Quick Notes	Shift + F7
Whiteboard	Shift + F8
File transcriber	Shift + F9
Audio editor	Shift + F10
AI Screenshot	PrtSc

All hotkeys are customisable in Settings. Per-prompt hotkeys are assigned per prompt via right-click → Assign hotkey.

---

AI Providers

Hotkeys works out of the box with no setup required, API access is built in.

If you want to use your own key (for higher limits or your own account), both providers are free:

Provider	Speed	Free tier	Sign up
Cerebras	~0.3 s	✅ Yes	cerebras.ai
Groq	~0.5 s	✅ Yes	console.groq.com

Sign up → copy your API key → paste it into Settings → done.

Voice-to-text (Whisper), file transcription, speaker diarization, and the whiteboard all run fully offline. No API key required for any of those features.

---

Privacy

• 🔒 Voice + file transcription run locally, Whisper model is bundled, nothing sent anywhere
• 🔒 Speaker diarization runs locally, pyannote model is bundled, runs in its own subprocess
• 🔒 Whiteboard runs locally, Excalidraw is bundled, no remote calls
• 🌐 Text refinement goes to Cerebras or Groq, same as any AI assistant you use
• 🚫 No analytics, no telemetry, no account required

---

Feature list

	Feature	Notes
✍️	AI text refiner	Works in any app, any text field
📚	Prompt library	16 built-in prompts, unlimited custom ones
⌨️	Per-prompt hotkeys	Assign F1-F10 to any prompt; fires from any app
🗒️	Sticky note popup	Preview and edit a prompt before it fires
🔤	Live spell check	Misspellings underlined in red as you type
🔍	Prompt search	Find prompts instantly as you type
🔄	Drag to reorder	Organise prompts by drag and drop
🗂️	Folders + colors	Group prompts into folders, color-code them
↩️	Undo last refine	Instantly revert an AI rewrite with Alt+Shift+Z
🎙️	Voice to text	Fully offline, Whisper bundled
🔇	Noise reduction	Works cleanly in noisy environments
🚀	Push-to-talk	Hold to record, release to transcribe
📋	Transcription history	Browse and copy past recordings
⚡	Instant paste	Output types directly where your cursor is
🖥️	System tray / menu bar	Zero UI clutter
🎨	Whiteboard	Offline Excalidraw, infinite canvas, all keyboard-driven (Shift+F8)
🎤	File transcriber	MP3/MP4/MKV/WebM/M4A/WAV/FLAC + YouTube URLs (Shift+F9)
👥	Speaker diarization	Speaker 1/2/... labels via pyannote, runs out-of-process for stability
🎵	Audio editor	Trim/mix/fade/export audio with bundled waveform editor (Shift+F10)
🔴	Macro recorder	Record & replay keystrokes + mouse clicks (Shift+F1)
💾	Saved macros	Name, save, and assign hotkeys to your most-used macros
🎬	Screen recorder	MP4, optional mic audio, any window or region (Shift+F2)
🎞️	GIF recorder	Animated GIF, any region, adjustable FPS (Shift+F3)
📸	AI screenshot	Capture + ask AI anything about it (PrtSc)
📝	Quick Notes	Floating scratchpad with image-paste OCR (Shift+F7)
🔗	Web search	Quick-launch web search with selection (Shift+F5)
⛓️	Prompt chains	Run several prompts in sequence on the same text (Shift+F6)
🔁	Hotkey watchdog	Auto-recovers if hotkeys stop responding
🍎	macOS support	Full feature parity on Mac

---

Running from source

git clone https://github.com/sprawf/hotkeys.git
cd hotkeys

# Windows
python -m venv venv
venv\Scripts\pip install -r requirements.txt
venv\Scripts\python main.py

# Mac
python3 -m venv venv
venv/bin/pip install -r requirements_mac.txt
venv/bin/python3 main.py

---

Tech stack

Component	Library
UI	CustomTkinter + tkinter
Speech-to-text	faster-whisper (offline)
Speaker diarization	pyannote.audio + torch (out-of-process worker)
Voice activity detection	Silero VAD
AI text refinement	Cerebras / Groq API
Spell check	pyspellchecker
Global hotkeys	keyboard
Macro recorder	pynput
Screen capture	win32ui + PyAV
Screen / GIF encoding	PyAV (FFmpeg)
System tray	pystray
Whiteboard	Excalidraw + pywebview (Edge WebView2)
Audio editor	Tenacity portable (bundled)
OCR	Groq vision API
Packaging	PyInstaller (onedir, double-spec for Hotkeys.exe + diarize.exe subprocess)

---

License

MIT, free to use, fork, and build on.

---

If this saved you time, consider leaving a ⭐, it helps others find it.