Download App
Home Documentation Language Settings

Language Settings

StarWhisper supports transcription and interface customization in multiple languages. Configure which language Whisper listens for and which language the app uses for its interface.

Two Independent Language Settings

StarWhisper has two separate language configurations that work independently:

Setting Purpose Default
Transcription Language Which language Whisper listens for when you speak Auto-detect
Interface Language The UI language for menus, buttons, and messages English

Transcription Language

The transcription language controls which language Whisper AI expects when processing your voice. This directly affects transcription accuracy.

Auto-Detect Mode (Recommended)

By default, StarWhisper uses auto-detect, which lets Whisper automatically identify the language being spoken. This works well for:

  • Single-language recordings (English, Spanish, French, etc.)
  • Users who occasionally switch between languages
  • General-purpose dictation where language may vary

When to Use Auto-Detect

Auto-detect works best for conversational speech in a single language. Whisper is highly accurate at identifying English, Spanish, French, German, and other major languages automatically.

Specific Language Mode

For improved accuracy, select a specific language when you know you'll be speaking primarily in that language. Benefits include:

  • Higher accuracy for technical terminology and accents
  • Faster processing since Whisper doesn't need to detect language first
  • Better handling of uncommon words and proper nouns

Supported Transcription Languages

Whisper supports transcription in 99+ languages. The most commonly used languages available in StarWhisper include:

  • English (en) - Best overall accuracy
  • Spanish (es) - Excellent support for all dialects
  • French (fr) - Strong performance
  • German (de) - High accuracy
  • Portuguese (pt) - Brazilian and European
  • Dutch (nl)
  • Russian (ru)
  • Chinese (zh) - Mandarin and other dialects
  • Japanese (ja)
  • Korean (ko)
  • Polish (pl)

For a complete list of supported languages, refer to the OpenAI Whisper documentation.

Interface Language

The interface language changes all UI elements, menus, tooltips, and messages in StarWhisper. This is completely separate from the transcription language.

Available Interface Languages

StarWhisper's interface is available in the following languages:

Language Code Coverage
English en 100% (Primary)
German de 100%
Spanish es 100%
French fr 100%
Portuguese pt 100%
Dutch nl 100%
Russian ru 100%
Chinese zh 100%
Japanese ja 100%
Korean ko 100%
Polish pl 100%

How to Change Language Settings

Opening Settings

  1. Right-click the StarWhisper recording circle
  2. Select Settings from the context menu
  3. Navigate to the Language section

Changing Transcription Language

  1. In Settings, find the Transcription Language dropdown
  2. Select your preferred language or Auto (detect language automatically)
  3. Changes take effect immediately for new recordings

Changing Interface Language

  1. In Settings, find the Interface Language dropdown
  2. Select your preferred UI language
  3. The interface will reload with the new language

Interface vs Transcription

Changing the interface language does NOT automatically change the transcription language. For example, you can have a German interface while transcribing English speech. Configure both settings separately based on your needs.

Multi-Language Dictation

Code-Switching (Mixed Languages)

Whisper has limited support for code-switching (mixing languages in a single recording). Performance varies:

  • Auto-detect: May detect the dominant language and transcribe accordingly
  • Specific language: Will try to force transcription in the selected language
  • Best practice: Record in one language at a time for optimal accuracy

Recommendations for Multi-Language Users

Scenario Recommendation
Occasional language switching Use auto-detect mode
Primary language + occasional English terms Select your primary language
Bilingual environment (e.g., Spanglish) Use auto-detect and accept reduced accuracy
Professional/technical work Select specific language for each recording

Tips for Best Results

Improving Transcription Accuracy

  • Match your use case: Auto-detect for casual use, specific language for professional work
  • Stick to one language per recording: Mixing languages reduces accuracy
  • Use larger models for non-English: The medium or large model performs better for languages other than English
  • Enunciate technical terms: Whisper may struggle with jargon in any language

Language-Specific Considerations

Language Tips
English Works exceptionally well on all models, even tiny
Spanish Excellent results with small or larger models
Chinese/Japanese/Korean Use medium or large for complex characters
German/French Handles compound words and accents well on small+
Less common languages Always use large model for best accuracy

Troubleshooting Language Issues

Incorrect Language Detection

Problem: Auto-detect chooses the wrong language

Solution:

  • Switch to a specific language setting instead of auto-detect
  • Use a larger model (medium or large) for better language identification
  • Speak more clearly or record longer phrases (5+ seconds)

Mixed-Language Transcription Errors

Problem: English words in non-English speech aren't transcribed correctly

Solution:

  • Select your primary language explicitly (not auto-detect)
  • Pronounce English terms with your native accent rather than an English accent
  • Consider using custom vocabulary for frequently used terms

Interface Language Not Changing

Problem: Interface remains in English after language change

Solution:

  • Close and reopen the Settings window
  • Restart StarWhisper completely
  • Check that you changed Interface Language (not Transcription Language)

Next Steps