Limitations & Constraints

VoxWrite is a powerful voice-to-text tool, but like any software, it has certain limitations. Understanding these constraints will help you use VoxWrite more effectively.

Browser & Platform Limitations

Chrome Extension Only

Current limitation:

• ✅ Works on: Chrome. Probably it works in all Chromium-based browsers (Edge, Brave, Opera), but it's still in a testing mode.
• ❌ Does NOT work on: Firefox, Safari, mobile browsers

Why: VoxWrite is built as a Chrome Extension using Chrome APIs.

Workaround: Use Chrome or Chromium-based browser for VoxWrite functionality.

Desktop Only (No Mobile Support)

Current limitation:

• ✅ Works on: Windows, Mac, Linux desktop
• ❌ Does NOT work on: iOS, Android, tablets

Why: Chrome Extensions don't run on mobile browsers.

Future: Mobile app may be considered based on user demand.

Internet Connection Required

Current limitation:

• AI transformation requires active internet connection
• Audio transcription requires connection to OpenAI/Anthropic servers

Why: AI models run on cloud servers, not locally.

Workaround: Basic transcription might work offline in future updates, but AI transformation always requires internet.

Website Compatibility Issues

Some Websites Block Microphone Access

Affected sites:

• Banking websites (security restrictions)
• Some enterprise applications
• Websites with strict Content Security Policy (CSP)
• Sites that override browser microphone permissions

Why: Website security policies prevent extensions from accessing microphone.

Workaround:

• Use VoxWrite in side panel to record, then copy/paste text
• Check if website allows microphone in browser settings

Text Insertion May Not Work Everywhere

Limitation:

• VoxWrite can't insert text into some protected fields:
• Password fields
• Credit card input fields
• Some rich text editors (WYSIWYG)
• Canvas-based text editors
• Shadow DOM elements

Why: Browser security prevents extensions from modifying certain input types.

Workaround: Copy text from VoxWrite and manually paste it.

Recording & Audio Limitations

Background Noise Affects Accuracy

Current limitation:

• High background noise reduces transcription accuracy
• Multiple speakers may confuse the AI
• Music or TV in background causes errors

Best practices:

• Record in quiet environment
• Use quality microphone
• Speak clearly and at moderate pace

AI & Language Processing Limitations

Language Support Varies by Model

Transcription (Audio Model):

• GPT-4o Mini Transcribe: 50+ languages
• Whisper-1: Excellent multilingual support
• Quality varies by language (English best supported)

Post-processing (AI Transformation):

• Works best with: English, Spanish, French, German, Chinese, Japanese
• May struggle with: Less common languages, dialects, slang

Limitation: Some languages may have lower accuracy or slower processing.

AI Transformation Not Perfect

Current limitations:

• AI may misinterpret context occasionally
• Complex instructions may not always be followed exactly
• Creative or nuanced writing may lose personal touch
• Technical jargon may be misunderstood

Why: AI models have inherent limitations in understanding context.

Best practice: Review and edit AI-transformed text before sending important messages.

Custom Prompts Have Limits

Constraints:

• Very complex prompts may confuse the AI
• Contradictory instructions cause unpredictable results
• Extremely long prompts may be truncated
• Some formatting requests may not work perfectly

Recommendation: Keep prompts clear, specific, and under 500 characters.

Performance Limitations

Processing Speed Varies

Factors affecting speed:

• AI model selected (faster models = lower quality)
• Recording length (longer = slower processing)
• Internet connection speed
• API provider server load
• Complexity of custom prompt

Typical processing times:

• 30 seconds of audio: 2-5 seconds to process
• 5 minutes of audio: 10-30 seconds to process

Feature Limitations

No Real-Time Transcription

Current limitation:

• You must stop recording before seeing transcribed text
• No live preview of transcription while speaking

Why: API calls happen after recording completes.

Future: Real-time transcription may be added in future updates.

No Collaboration Features

Current limitation:

• No team sharing of recordings
• No shared custom rules or templates
• No multi-user accounts

Why: VoxWrite is designed as individual productivity tool.

Future: Team features may be considered based on demand.

No Offline Mode

Current limitation:

• Cannot transcribe audio without internet
• Cannot apply AI transformations offline
• Extension requires online activation

Why: AI models run on cloud servers.

Partial workaround: Record audio offline, process when connection returns (not yet implemented).

No Video or Screen Recording

Current limitation:

• Audio only - no video recording
• No screen capture integration
• No meeting recording (use dedicated tools)

Why: VoxWrite focuses on voice-to-text, not multimedia recording.

Conclusion

VoxWrite is actively developed and many limitations will be addressed in future updates. Understanding current constraints helps you:

• Set realistic expectations
• Find workarounds when needed
• Use VoxWrite more effectively
• Provide valuable feedback for improvements

Most limitations don't affect daily use for 90% of users. The core functionality—voice to text with AI transformation—works reliably across most websites and use cases.

If you have any questions, feature request, whatever, please use our contact form.