The Gemma 4 era has arrived on iOS! Silent Scribe brings full support for the powerful Gemma4-e2b and Gemma4-e4b models. And the best part? It’s not just for the newest hardware. We’ve highly optimized performance so your trusty older devices (including iPhone 12, 13, 14, and 15 Pro/Pro Max) can run Gemma 4 smoothly and flawlessly. Are you worried about data privacy leaks with cloud-based AI? Tired of emptying your wallet on expensive API tokens? Say goodbye to those limitations. Experience the latest open-source AI models right in your pocket. Crush your tasks anytime, anywhere—100% offline, with absolute privacy and zero recurring costs.

Bug Fixes

- Flexible AI Model Management: You now have the freedom to choose! Easily browse, download, and manage a selection of advanced Large Language Models (LLMs) directly within the app. Download the models that fit your current needs, switch between them effortlessly, and delete them anytime to free up storage space. - Optimized Performance Across Devices: We’ve significantly improved memory efficiency for running on-device AI. Whether you're using the newest iPhone to explore a wider variety of models, or using an older device to experience advanced local AI for the first time, you'll enjoy a smoother, more stable experience. - Enhanced Offline Productivity: Handle your text processing and daily tasks faster and more conveniently than ever before. Everything runs 100% offline, keeping your workflow completely private and secure.

[New] GPU Acceleration: Transcription for both live and local audio now supports GPU acceleration. [Pro Tip] If your device is compatible, turn on the GPU option in Settings for a massive speed improvement! [Fixes] We've fixed some known bugs and made general optimizations to resource usage and performance.

[New] Added keyboard shortcuts support for iPad. [Fix] Bug fixes and performance improvements.

[New] Expanded Language Support: Added support for Dutch and Korean. AI Smart Summary: The new summarization feature now supports audio files up to 1 hour, helping you quickly capture key insights. [Improved] Deep Improved Optimization: Data models are now fine-tuned for specific devices, ensuring smoother performance and lower resource consumption. [Fix] Fixed several known issues (and probably created a new one).

[New] System-Wide Share Extension: Now you can import media directly from other apps! Just use the share menu in apps like Files, Voice Memos, Photos, or WhatsApp to send files straight to our app for processing. [New] Home Screen Quick Actions: Need to transcribe something fast? Long-press our app icon on your Home Screen and tap "Live Transcription" to start recording instantly. [Improved] Smarter & Faster Imports with VAD: We've added Voice Activity Detection (VAD) to our import feature, making your transcriptions faster and more accurate. [Improved] Faster AI Performance: We've optimized our AI engine, so all AI tasks now run faster. [Fix] Fixed several known issues and improved overall app stability.

[New Feature] On-Device AI Model: Deployed a completely new on-device AI model. [New Feature] New AI Tasks: Offline Translation and Grammar are now supported. [New Feature] New Language Support: Added support for Portuguese and Spanish (Mexico). [New Feature] FLAC Format Support: You can now import and process FLAC lossless audio files. [Enhancement] Speech Recognition: Improved the accuracy and robustness of the speech recognition engine. [Enhancement] Compatibility & UI: Updated to the new iOS 26 Liquid Glass design, with adjustments to various UI elements and text. [Fix] Fixed several known issues and improved overall app stability.

The Speed & Stability Update This update is all about speed. We've optimized code, squashed bugs, and fine-tuned performance from top to bottom. The app should now feel faster, smoother, and more reliable than ever. Thanks for using Silent Scribe

* Core speech recognition improvements + Added a device environment security check

+ Real-Time Audio Visualization: Dynamic audio waveform display now integrated into the live transcription interface, providing immediate visual feedback on input levels. + Cache Management Utility: Introduced a new option for users to manually clear the application cache, aiding in storage optimization and maintaining peak performance. * Stability Enhancements & Bug Resolution: This update includes several under-the-hood bug fixes and performance optimizations to improve overall application stability and user experience.

* Improvement: The progress bar now displays a visual waveform during audio playback.

+ Shiny New Welcome Mat: Our onboarding pages got a makeover. Looking sharp, if we do say so ourselves! + More File Friends: We now play nice with MKV & WEBM files. Bring on the obscure formats, we dare ya! + Audio Control Freak Mode (Activated!): Importing media? Now you can pick audio tracks, tell background noise to take a hike (noise reduction!), crank up the volume, AND make voices clearer than your conscience after finally doing the dishes. + Control Center: Adds "Now Playing" info and controls during audio playback. + Konnichiwa, Japan: That's right, we've added Japanese transcription! Our AI is now officially bilingual (well, multilingual, but Japanese is the new star!). + iPhone Facelift: The UI got some love. It’s like your favorite app, just… snazzier. + AI Got Faster (Again): Our AI transcription engine just chugged an espresso. Expect even speedier results. + Smarty Pants AI Scaling: Got an older iPhone? Or a beastly new one? Our AI now flexes its model size to match your device's muscle. We even give ya the performance numbers, 'cause we're nerds like that. + Grab Your AI Models: New AI model resource downloads are here for peak performance!

+ Supports importing media files (M4A/AAC).

+ Supports importing local media files (MOV, MP4 (H.264/H.265), MP3, WAV). + Supports rich text editing (font size, text color, bold, italic, underline). + Supports exporting and sharing notes (as audio files, PDF). + Supports marking time segments in extracted text.