AI Scribe

AI Scribe is a privacy-first, 100% free utility designed to handle audio transcription and Image-to-Text (OCR) without subscriptions or hidden costs. It bridges the gap between high-speed cloud processing and total on-device privacy.

The Hybrid Edge:

Local Mode: Runs Google’s Gemma 3n E2B multimodal model entirely offline via LiteRT. Your voice notes and photos never leave your device. It uses a custom Sequential Chunking Algorithm to process long audio files in 30-second segments, overcoming mobile RAM limits and model context degradation.

Cloud Mode (BYO-Key): For older devices or maximum speed, use the Gemini API. Simply plug in your own free Google API Key.

Privacy Architecture: Unlike other apps, AI Scribe has no backend server.

Your API Key is stored exclusively in the app's Internal Sandbox (private local storage).

Data is transmitted directly from your device to Google’s servers (Cloud) or processed 100% offline (Local).

No tracking, no ads, no middleman.

Key Features:

WhatsApp Integration: Share any voice note directly to AI Scribe for instant text + auto-summary.

Multimodal OCR: Extract text from screenshots or documents using the same local/cloud logic.

Multi-Language Support: Optimized prompting for English, Italian, French, German, Spanish, and Portuguese.

Vibe Coding Origin: Built through a synergy of Human Architecture and AI Agents (Claude Opus/Gemini Pro), specifically engineered to bypass the technical hurdles of on-device multimodal ingestion.