gyute.fyi Yap home
Configuration

Using Clanker Yap

Configuration

4 min read

Detailed explanation of all Clanker Yap configuration options.

Configuration Storage

Settings are stored in a SQLite database at:

  • Linux: ~/.local/share/voice-transcribe/voice-transcribe.db

Settings Schema

FieldTypeDefaultDescription
hotkeyStringCmdOrCtrl+Shift+VGlobal push-to-talk hotkey
model_pathString~/.local/share/voice-transcribe/ggml-base.en.binPath to Whisper model
model_nameStringggml-base.en.binDisplay name for the model
paste_modeStringautoClipboard paste behavior (auto, standard, terminal)
auto_pasteBooleantrueWhether to automatically simulate Ctrl+V after copying to clipboard
audio_inputJSONnullSelected microphone device
schema_versionInteger1Settings format version

Hotkey Configuration

Format

Hotkeys use Tauri accelerator format: <Modifier>+<Key>

Valid modifiers:

  • CmdOrCtrl - Control on Linux
  • Ctrl - Control key
  • Alt - Alt key
  • Shift - Shift key
  • Super - Super key

Requirements:

  • At least one modifier key is required
  • Must include one non-modifier key

Examples

HotkeyDescription
CmdOrCtrl+Shift+VDefault (Ctrl+Shift+V on Linux)
Alt+RAlt+R
CmdOrCtrl+Alt+SpaceAll modifiers plus space

Setting via UI

  1. Click the Hotkey toolbar button
  2. Click "Record shortcut"
  3. Press your desired key combination
  4. Click "Save"

Model Configuration

Supported Formats

Clanker Yap uses GGML-format Whisper models. These are .bin files containing quantized weights.

ModelSizeSpeedAccuracyBest For
ggml-base.en.bin148 MBFastestGoodEnglish-only, fast transcription
ggml-small.en.bin466 MBFastBetterEnglish-only, higher accuracy
ggml-medium.en.bin1.5 GBMediumVery GoodEnglish-only, detailed capture
ggml-base.bin148 MBFastestGoodMultilingual, base quality
ggml-small.bin466 MBFastBetterMultilingual, balanced

Setting Custom Path

  1. Click the Model toolbar button
  2. Click "Edit"
  3. Enter full path (e.g., /home/user/models/whisper-base.bin)
  4. Click "Save"

Downloading Models

Via UI:

  1. Click "Download base.en" in the Model settings
  2. Wait for download to complete
  3. Model is saved to default location

Manually:

curl -L "MODEL_URL" -o ~/.local/share/voice-transcribe/MODEL_NAME.bin

Model URLs from ggml-org/whisper.cpp on HuggingFace.

Audio Input Configuration

Selection Types

{
  "type": "system_default"  // Use system default device
}
{
  "type": "by_name",
  "value": "USB Microphone: Audio (hw:1,0)"
}

Setting via UI

  1. Click the Mic toolbar button
  2. Select device from dropdown
  3. Click "Save"

Finding Device Names

List available devices via the microphone dropdown in the UI, or check:

  • Linux: arecord -l or pactl list sources

Paste Mode Configuration

Controls how transcribed text is inserted into applications.

Auto-paste

The auto_paste setting controls whether Clanker Yap automatically simulates a Ctrl+V (or Cmd+V) keystroke after copying the transcription to the clipboard.

  • Enabled (default): Text is copied to the clipboard and automatically pasted.
  • Disabled: Text is only copied to the clipboard — press Ctrl+V manually.

On KDE/Wayland, the keyboard simulation uses a persistent input controller that is initialised once per session. You may see a single "Remote Control" permission prompt the first time auto-paste runs. Subsequent pastes reuse the same session without additional prompts.

Paste Shortcut Style

Modes

ModeBehaviorUse Case
autoStandard paste, fallback to keyboard simulationDefault, works in most apps
standardDirect clipboard paste via TauriMost applications
terminalType character-by-characterTerminals, Vim, applications that block paste shortcuts

Comparison

AspectStandardTerminal
SpeedFastSlow
UnicodeSupportedLimited
Special charsHandledMay have issues
ClipboardModifiedNot modified

Setting via UI

  1. Click the Paste toolbar button
  2. Select mode (Auto/Standard/Terminal)
  3. Click "Save"

Configuration Files

tauri.conf.json

Main Tauri configuration file located at src-tauri/tauri.conf.json:

{
  "productName": "Clanker Yap",
  "version": "X.Y.Z",
  "identifier": "dev.jay.voice-transcribe",
  "app": {
    "windows": [{
      "width": 480,
      "height": 196,
      "resizable": false
    }]
  }
}

Key options:

  • productName - Window title and app name
  • version - App version (used in builds)
  • identifier - Unique app identifier (required for stores)
  • width/height - Window dimensions
  • resizable - Whether window can be resized

Customizing Window Size

Edit src-tauri/tauri.conf.json:

{
  "app": {
    "windows": [{
      "label": "main",
      "width": 600,
      "height": 400
    }]
  }
}

Environment Variables

No environment variables are required. All configuration is stored in the SQLite database.

Resetting to Defaults

To reset all settings:

  1. Delete the database file:
    rm ~/.local/share/voice-transcribe/voice-transcribe.db
    
  2. Restart the application

This will recreate the database with default values.