Privacy Policy
Script to Speech is committed to protecting your privacy and data. This document explains what data is collected, how it’s used, and what you should know about third-party services when using this tool.
Data Collection by Script to Speech
Script to Speech collects NO user data. Specifically:
- No telemetry or analytics: The application does not send usage statistics, error reports, or any other data about your usage
- No tracking: No user behavior is monitored or recorded
- No advertisements: No ads are displayed or user data sold to advertisers
- No data sharing: Script to Speech does not share any user data with third parties
- No network requests: The application only makes network requests to services required for functionality that you explicitly configure
- No remote logging: All logs are stored locally on your machine
- No user accounts: No registration, login, or user account system exists
Local Data Storage
Script to Speech stores data locally on your machine:
- Input files: Screenplay PDFs/TXT files are copied to
input/[screenplay_name]/directory - Generated files: Parsed JSON, voice configurations, and audio files are stored in
input/andoutput/directories - Cache files: Audio clips are cached locally in
output/[screenplay_name]/cache/ - Log files: Processing logs are stored in
output/[screenplay_name]/logs/
You have full control over all local data and can delete any files at any time.
Data Sent to External Services for Functionality
Script to Speech requires external services to provide its core functionality. Here’s what data is sent and why:
Audio Generation (Required for TTS Functionality)
To convert your screenplay text into speech, Script to Speech sends individual dialogue chunks to TTS (text-to-speech) providers you configure. Support for local, self-hosted, TTS providers is on teh Script to Speech roadmap.
TTS Providers that may receive your content, if configured:
- OpenAI
- ElevenLabs
- Cartesia
- Minimax
- Zonos
Data sent: Individual dialogue chunks (typically single lines of dialogue or scene descriptions) are sent one at a time to generate audio. Your screenplay is not sent as a complete document to TTS providers.
Character Notes Creation (Optional Feature)
If you choose to use the LLM-assisted character note creation feature (sts-generate-character-notes-prompt), this is entirely optional and can be skipped.
LLM Services (when you choose to use voice casting):
Data sent:
- Your complete screenplay text
- Your current voice configuration
- Instructions for voice casting analysis
Important: This optional feature sends your ENTIRE screenplay to the LLM service. Only use this feature with LLM providers whose privacy policies you trust, or skip this feature entirely and configure voices manually.
Voice Library Casting (Optional Feature)
If you choose to use the LLM-assisted voice library casting feature (sts-generate-voice-library-casting-prompt), this is entirely optional and can be skipped.
LLM Services (when you choose to use voice library casting):
Data sent:
- Your voice configuration including character names and any casting notes
- Voice library data for the specified providers
- Instructions for voice selection
Important: This optional feature sends your character list and casting notes (but NOT your screenplay text) to the LLM service. This is more privacy-conscious than character notes generation, but still shares information about your project’s characters.
Privacy-conscious approach: For maximum privacy, manually configure voices without using either LLM-assisted feature. For moderate privacy, manually add casting notes to your configuration, then use only voice library casting (avoiding screenplay text sharing).
What You Should Know About Third-Party Services
Each service has different policies regarding:
- Data retention: How long they keep your content
- Data usage: Whether your content is used for training AI models
- Data sharing: Whether your content is shared with other parties
- Geographic storage: Where your data is processed and stored
Recommended Actions
Before using any service:
- Read their privacy policy: Understand how your data will be handled
- Check training data policies: Determine if your content will be used to train AI models
- Review data retention: Understand how long your content is stored
- Consider data sensitivity: Evaluate whether you’re comfortable sharing your screenplay content
Service-Specific Resources
- OpenAI: Privacy Policy | Consumer Privacy / Data use
- ElevenLabs: Privacy Policy
- Cartesia: Privacy Policy
- Minimax: Privacy Policy
- Zonos: Privacy Policy
Recommendations for Privacy-Conscious Usage
Minimize Data Exposure
- Test with sample content: Use non-sensitive text for initial testing and voice sampling
- Use dummy providers: Test configurations with
--dummy-tts-provider-overrideflag
Read Privacy / Data Usage Policies
- Different TTS and LLM providers have different stances on data privacy and usage. Understand which providers use user supplied content for LLM training
- “Free” tiers for providers often give up privacy and data usage rights in exchange for free usage
LLM Voice Casting Considerations (Optional Features)
- Manual approach: Skip LLM assistance entirely and configure voices manually - these features are completely optional
- Use local LLMs: Consider running local language models instead of cloud services
- Custom prompts: Create your own voice casting prompts without including screenplay text
- Avoid sensitive content: Don’t use voice casting features for confidential screenplays
API Key Security
- Use .env files: Store API keys locally rather than in environment variables
- Rotate keys regularly: Generate new API keys periodically
- Limit key permissions: Use API keys with minimal required permissions where possible
- Monitor usage: Check provider dashboards for unexpected API usage
Data Management
- Clean up regularly: Delete cached audio and logs you no longer need
- Backup configurations: Keep voice configurations but consider removing screenplay content from backups
- Secure storage: Store sensitive screenplay files in encrypted directories
- Version control: Avoid committing screenplay content or API keys to version control
International Users
Different countries have different privacy laws (GDPR, CCPA, etc.). Consider:
- Data residency: Where your data is processed by third-party services
- Legal compliance: Whether service providers comply with your local privacy laws
- Data transfer: How data moves between countries when using cloud services
Changes to This Policy
This privacy policy may be updated to reflect changes in:
- How Script to Speech operates
- New third-party service integrations
- Legal requirements
- User feedback
Check this file periodically for updates. The last update date is shown at the bottom of this document.
Contact Information
For privacy-related questions or concerns about Script to Speech:
- GitHub Issues: Create an issue for privacy questions
- Email: Contact the maintainer at the email address listed in the project’s
pyproject.toml
For questions about third-party service privacy policies, contact those services directly using the links provided above.
Your Responsibility
As a user of Script to Speech, you are responsible for:
- Understanding the privacy policies of services you choose to use
- Making informed decisions about what content to process
- Complying with any applicable laws regarding data processing
- Protecting your API keys and sensitive content
Summary
- Script to Speech: Collects no data, operates locally, fully under your control
- TTS Providers: Receive individual text lines for audio generation
- LLM Services: May receive complete screenplay content when using voice casting features
- Your choice: You decide which services to use and what content to process
Last updated: June 25, 2025