Chapter 12: Preferences, License Plans & FAQ

12.1 Preferences & Settings

Configure Scribis in "Settings":

SCREENSHOT BLUEPRINT

Scribis Settings page showing language selection, model cache management, and API key configurations

Language & Appearance: Toggle between English and Traditional Chinese; select Dark, Light, or System theme.
Model Management: Download, update, or purge local ASR models and TTS weights to manage disk space.
API Keys (BYOK): Add OpenAI, Gemini, Claude, OpenRouter keys, or connect a local Ollama endpoint.
Global Shortcuts: Customize hotkeys for Dictation and Editor controls.

12.2 License Tiers & Feature Comparison (Free vs. Pro)

Scribis offers Free and Pro license tiers:

SCREENSHOT BLUEPRINT

License management view displaying Pro subscription status, active device seats, and offline certificate status

💎 Feature & Limit Comparison Table:

Feature / Limit	Free Tier	Pro Tier
File Transcription Length	⏱️ Max 30 minutes per file	🚀 Unlimited file duration
Live Recording / Dictation	⏱️ Max 10 minutes per session	🚀 Unlimited recording time
Custom & Correction Dictionaries	❌ Not supported in Free	✅ Full unlimited dictionaries & sync
Batch Upload Queue	❌ Not supported in Free	✅ Unlimited batch upload queue
Voiceprint Speaker Library	Basic trial (max 2 profiles)	✅ Unlimited profiles & auto-groups
AI Assistant & GEC Correction	Basic trial quota	✅ Full access to advanced AI tools
MCP External AI Tools	Basic tools	✅ Full access to all editing & vision tools
Offline License Support	Online verification mode	✅ 30-day lease + permanent offline bound license

🔒 Offline Leases & Device Seats:

30-Day Offline Grace Period: Pro users can work completely offline for up to 30 days after launching with an active internet connection.
Device Seat Management: Each Pro license supports 2 concurrent devices. Deactivated seats enter a 7-day security cooling period.
Permanent Offline Bound License: After 30 days of active Pro ownership, you can generate a machine-locked permanent offline certificate for lifelong internet-free usage!

12.3 Frequently Asked Questions (FAQ)

Q1: Does Scribis upload my audio recordings to the cloud?

A: No. Scribis operates under a strict "Local-First" privacy model. All ASR transcription and video rendering take place locally on your computer's CPU/GPU. Audio is never sent to external servers unless you explicitly call third-party cloud LLM APIs.

Q2: What causes repeated words or phantom phrases like "Thank you for watching"?

A: This is a known behavior of speech models when encountering long silent gaps or background noise.

Solution 1: Enter a brief Prompt in the transcription settings to guide the model.

Solution 2: Use Alt + D to quickly delete phantom lines in the editor, or use Chapter 6: GEC Correction to clean them up.

Q3: How can I achieve the highest possible transcription accuracy?

A:

For English audio, choose Parakeet v3 TDT (#1 ranked for English).

For high-stakes or multilingual recordings, choose VibeVoice.

Manually select the specific language (e.g., English or Chinese) instead of leaving it on Auto.

Q4: Can I run Scribis on a laptop without a dedicated GPU?

A: Absolutely! Choose the SenseVoice Small model. It is optimized for CPU inference and runs at 10x to 15x real-time speed even on ultra-thin productivity laptops.

Q5: How do I contact support?

A: For business licensing, feedback, or support inquiries, contact our team at: [email protected].