Frequently Asked Questions

Simply click the purple circle to start recording, speak your text, then click again to stop. Your transcription will automatically be copied to clipboard and pasted into your active window. You can also use the CapsLock hotkey to record without clicking.

No! StarWhisper works completely offline using local AI models. Your voice data never leaves your computer, ensuring complete privacy. However, Pro users can optionally use OpenAI's cloud API if preferred.

Free Plan (5,000 words/week):

• Tiny, Base & Small AI models
• Local offline transcription
• 3 visual themes
• GPU acceleration

Pro Plan ($10/month - Unlimited):

• Unlimited transcriptions
• Medium & Large models (higher accuracy)
• OpenAI API access (cloud transcription)
• File transcription (WAV, MP3)
• Premium themes
• Priority support

Right-click the circle → Settings → Basic tab → Hotkey section. You can set any key combination. Default is CapsLock.

GPU acceleration uses your NVIDIA graphics card to transcribe 5-10x faster than CPU-only mode. If you have an NVIDIA GPU, we highly recommend downloading the GPU pack (~528MB). The app will prompt you to download it automatically if a compatible GPU is detected.

Model quality vs speed:

• Tiny (75MB) - Fastest, good for simple dictation
• Base (142MB) - Balanced speed and accuracy
• Small (466MB) - Recommended for most users
• Medium (1.5GB, Pro) - Higher accuracy
• Large (2.9GB, Pro) - Best accuracy, slower

We recommend Small for the best balance of speed and accuracy.

StarWhisper supports 29+ languages including English, Spanish, Chinese, Russian, Japanese, French, German, Arabic, Hindi, Tagalog, and many more. You can set your preferred language in Settings → Transcription → Language, or use Auto-detect.

Absolutely! StarWhisper works with any application on Windows. Many users use it for dictating to ChatGPT, Claude, coding, writing documents, emails, and more. The auto-paste feature makes it seamless.

StarWhisper uses OpenAI's Whisper model, which achieves human-level accuracy for English and high accuracy for other major languages. Accuracy depends on:

• Model size (larger = more accurate)
• Audio quality (clear speech works best)
• Language (English has highest accuracy)
• Background noise (quiet environments recommended)

No! By default, all transcription happens locally on your computer. Your voice recordings never leave your device. We are privacy-first by design. Pro users can optionally enable OpenAI cloud transcription if they choose.

We never upload your transcriptions to our servers. Everything is stored locally on your computer in %AppData%\StarWhisper\history.json. You have full control over your data and can delete it anytime from the History tab.

We collect minimal data:

• Account email (if you sign in)
• Subscription status (Free or Pro)
• App version (for updates)

We do NOT collect:

• Your transcriptions or voice recordings
• Usage statistics or analytics
• Any personal data beyond your email

Common solutions:

• Check your microphone is working (test in Windows Settings)
• Grant microphone permission when prompted
• Select correct microphone in Settings → Audio tab
• Ensure you have a Whisper model downloaded (check Transcription tab)
• Try speaking louder or in a quieter environment

Enable GPU acceleration! If you have an NVIDIA graphics card:

• Open Settings → Transcription tab
• Check if "NVIDIA GPU detected" message appears
• Download GPU Pack (~528MB) if prompted
• Enable "Use GPU Acceleration" checkbox

This makes transcription 5-10x faster. Alternatively, use a smaller model (Tiny or Base) for faster processing.

Free plan users have a 5,000 word weekly limit. When reached, you'll see an error message. Options:

• Upgrade to Pro for unlimited transcriptions ($10/month)
• Wait until next week when the limit resets (every 7 days)

Your usage resets automatically each week. Check your current usage in Settings → Plan tab.

Tips for better accuracy:

• Use a larger model (Small, Medium, or Large for Pro users)
• Set specific language instead of auto-detect (Settings → Transcription)
• Speak clearly and at moderate pace
• Use a good quality microphone
• Reduce background noise
• Enable GPU acceleration for faster, more stable processing

Uninstall via Windows Settings → Apps → StarWhisper → Uninstall. The uninstaller will:

• Remove: Application files, GPU pack binaries, logs
• Keep: Your settings, account login, transcription history

This means you can reinstall later without losing your configuration. To completely remove all data, manually delete %AppData%\StarWhisper after uninstalling.

Click the Upgrade button in the app (red button in sidebar) or visit starwhisper.ai. Payment is processed securely via Stripe. Your Pro features activate immediately after payment.

Yes, anytime! Open Settings → Plan → Manage Subscription to cancel. You'll retain Pro features until the end of your current billing period, then automatically revert to the Free plan.

We accept all major credit cards (Visa, Mastercard, American Express, Discover) via Stripe. Payment is secure and encrypted.

Yes, we offer a 7-day money-back guarantee. If you're not satisfied with Pro, contact us at support@regios.org within 7 days of purchase for a full refund.

Pro users can transcribe entire audio files (WAV, MP3) instead of just live voice. Perfect for:

• Transcribing recorded interviews or meetings
• Converting podcasts to text
• Processing lecture recordings

Access via Settings → Files tab. Select your audio file and get a complete transcription that you can copy or download as text.

Yes! StarWhisper includes 8 built-in themes (Classic, Twilight, Electric, Sunset, Ocean, Emerald, Magenta, Lime). Free users get 3 themes, Pro users get all 8 plus a Custom theme editor where you can create your own color schemes.

Enable wake words in Settings → Audio tab. Choose from "Alexa", "Hey Mycroft", "Hey Jarvis", or "Computer". Once enabled, just say the wake word and StarWhisper will start recording automatically. Say "end transcript" to stop recording.

Minimum:

• Windows 10 or later (64-bit)
• 4GB RAM
• 500MB disk space (+ models)
• Working microphone

Recommended for GPU acceleration:

• NVIDIA GPU (GTX 10 series or newer)
• 8GB RAM
• 2GB disk space (includes GPU pack)

StarWhisper checks for updates automatically. When an update is available, you'll see a notification in Settings → About tab. Click "Download Update" to get the latest version. Updates are typically small (~189MB) and install automatically.

No! Your GPU pack, settings, transcription history, and account login all persist across updates. You won't need to re-download anything or reconfigure the app.

• In-app: Settings → Feedback tab (with optional diagnostic logs)
• Email: support@regios.org
• Contact form: starwhisper.ai/contact

We typically respond within 24 hours (faster for Pro users).