Skip to main content
SlasshyWispr is a desktop voice dictation and AI assistant app. This guide will help you set up the application and perform your first voice dictation.
SlasshyWispr is in strong beta. Expect bugs, rough edges, and breaking changes while features are actively evolving.

​
Prerequisites

Before you begin, ensure you have:

Windows OS

SlasshyWispr currently supports Windows operating systems

Microphone

A working microphone for voice input

Internet Connection

Required for online models and downloads

API Keys (Optional)

For online STT/AI models (OpenAI, etc.)

​
Installation

1

Download the Installer

Get the latest Windows installer from the GitHub releases page:https://github.com/SlasshyOverhere/SlasshyWispr/releases/latest
2

Run the Installer

Execute the downloaded installer and follow the on-screen instructions to complete the installation.
3

Launch SlasshyWispr

After installation, launch the application from your Start menu or desktop shortcut.

​
Initial Setup

​
Configure Models

1

Open Model Settings

Navigate to Settings > Models in the application.
2

Choose STT Runtime Mode

Select your Speech-to-Text (STT) runtime mode:
  • Online: Use cloud-based STT services (requires API key)
  • Offline: Use local STT models (no internet required)
3

Choose AI Runtime Mode

Select your AI assistant runtime mode:
  • Online: Use cloud-based AI models (requires API key)
  • Offline: Use local Ollama models

​
Configure Online Models (If Using Online Mode)

If you selected online mode for either STT or AI:
1

Add API Configuration

In the model settings:
  • Enter your API Base URL
  • Enter your API Key
  • Specify the STT Model Name (if using online STT)
  • Specify the AI Model Name (if using online AI)
2

Save API Credentials

Enable “Remember API Key” to avoid re-entering credentials on each launch.
Your API credentials are stored locally and never shared with third parties.

​
Configure Local Models (If Using Offline Mode)

​
Local STT Setup

If you selected offline STT mode:
1

Download a Local STT Model

In Settings > Models, download one of the available local STT models:
  • Parakeet v3 (478 MB) - Recommended for most users
  • Whisper Turbo (1.6 GB) - High accuracy
  • Moonshine Base (58 MB) - Lightweight option
2

Select the Model

Once downloaded, select the model from the dropdown to activate it.
The Parakeet v3 model offers the best balance of speed and accuracy for most use cases.

​
Local AI Setup

If you selected offline AI mode:
1

Configure Ollama

Ensure Ollama is installed and running on your system.Default Ollama URL: http://127.0.0.1:11434
2

Pull an Ollama Model

In the settings, select or pull a local Ollama model (e.g., llama2, mistral, phi3).
Learn more about Ollama models at https://ollama.com/library

​
Configure Microphone and Hotkeys

1

Open General Settings

Navigate to Settings > General.
2

Select Microphone

Choose your preferred microphone device from the dropdown list.
3

Configure Hotkeys

Review and customize your hotkeys:
  • Push-to-Talk Hotkey: Ctrl+Space (default) - Hold to record dictation
  • Command Hotkey: Ctrl+Shift+Space (default) - Trigger assistant mode
4

Choose Capture Mode

Select your preferred capture mode:
  • Push-to-Talk (default): Hold the hotkey while speaking
  • Single-Tap: Press once to start, press again to stop

​
Your First Dictation

Now you’re ready to test voice dictation!
1

Position Your Cursor

Click into any text field (e.g., Notepad, Word, browser text box).
2

Hold the Hotkey

Press and hold Ctrl+Space (or your configured hotkey).
3

Speak Clearly

While holding the hotkey, speak naturally: “Hello, this is my first dictation test.”
4

Release to Process

Release the hotkey when you’re done speaking. SlasshyWispr will process your voice and transcribe it.
5

View Results

Your transcribed text will be automatically pasted into the text field (if auto-paste is enabled) or copied to your clipboard.
Watch the app window for real-time pipeline status showing STT, AI, and TTS processing times.

​
Test Assistant Mode

SlasshyWispr includes an AI assistant mode for interactive conversations:
1

Enable Wake Word (Optional)

In Settings > General, enable “Wake Word” and set your assistant name (default: “Lily”).When enabled, say “Hey [Assistant Name]” to activate the assistant.
2

Trigger Assistant Mode

Use the command hotkey (Ctrl+Shift+Space by default) or say your wake phrase.
3

Ask a Question

Speak your question or command: “What’s the weather like today?”
4

Listen to Response

The assistant will generate a response and speak it back to you using text-to-speech (TTS).
Assistant mode requires an AI model (online or offline) to be configured.

​
Next Steps

Congratulations! You’ve completed your first voice dictation. Here are some advanced features to explore:

Dictionary & Snippets

Create custom vocabulary and text shortcuts

Style Profiles

Customize AI response tone and formatting

Context Awareness

Enable selection-based rewriting and context

TTS Configuration

Fine-tune text-to-speech voice and quality

​
Additional Resources

Need help? Open an issue on GitHub or check the documentation for detailed configuration guides.

Build docs developers (and LLMs) love

Get started for freeTalk to us