Binot v1.0.4

Release Notes v1.0.4

• Added support for rendering mathematical formulas and symbols with KaTeX
• Improved AI processing to strictly enforce language translation on Tidy and Summarize modes.

Binot v1.0.3

What's New in v1.0.3

We've brought massive upgrades to Binot to make your voice-to-text and summarization experience faster and smarter!

Key Features

• Multi-AI Engine Support: You can now choose between Google Gemini and Groq (Llama 3)!
  • Gemini: The beast for handling massive audio files (hours long) and deep summarization.
  • Groq (Recommended): Blazing fast processing speed with a generous rate limit.
• Dynamic Configuration: Seamlessly switch between AI engines in Onboarding and Settings. Your API Keys are securely saved and automatically swapped based on your preference.
• Flawless Mathematical Formatting: AI engines (especially Groq) are now strictly trained to accurately format mathematical symbols (e.g., ², ³, °, ±, ½) and prevent messy Markdown code blocks. Your text will always look perfectly elegant!

Improvements & Fixes

• Simplified API Key instructions in the Settings and Onboarding screens for a cleaner UI.
• Added smart audio limit detection: The app will now politely warn you if your audio exceeds Groq's 25MB limit and suggest switching to Gemini for long recordings.
• Various UI tweaks for a more balanced and proportional layout.

Binot v1.0.1

Warning

Backup your data before updating. This update includes changes to the app database schema. If you experience a force close after updating, clearing the app data will resolve the issue.

v1.0.1

New Features

Recording Modes

Two recording modes are now available in Settings:

• Fast Mode — Uses standard Speech-to-Text for real-time transcription. Lightweight and battery-friendly, no audio file stored.
• Accurate Mode — Records long sessions (hours) by saving audio and processing it with Gemini AI when the note is opened. Produces cleaner punctuation, grammar, and technical/math terminology. Live transcription is disabled in this mode.

Transcript Editing

Transcription results can now be manually edited directly within the app.

Text Highlight

Selected text in transcription results can be highlighted and annotated with custom notes.

AI Explain

Select any term or concept in a transcription to get an instant explanation powered by Gemini, without leaving the app.

---

Improvements

• Gemini File API — Accurate Mode now uses Gemini File API instead of inline Base64 encoding, removing RAM limitations for long recordings.
• Waveform Sensitivity — Amplitude visualization in Accurate Mode now uses logarithmic dB calculation at 33 FPS for more responsive feedback on quiet audio.
• Recording Guard — Switching recording modes mid-session now triggers a warning to prevent accidental data loss.
• Processing Indicator — Upload, processing, and transcribing stages are now shown as distinct steps while Gemini processes audio.

Binot v1.0.0

Binot v1.0.0 - Initial Release

This is the first stable release of Binot. Binot is a native Android application designed to accurately transcribe spoken audio and intelligently process it into structured, readable markdown notes using the Google Gemini API.

Built with a local-first philosophy, this application requires users to provide their own Gemini API key, ensuring absolute data privacy and personal autonomy over AI usage limits.

Features Included in This Release:

• Live Audio Transcription: Real-time voice dictation featuring an automated hardware-volume override to prevent notification sounds from interrupting the recording.
• AI Text Processing: Capabilities to tidy up syntax, summarize lengthy dictations, and translate transcripts across multiple languages using the Gemini 2.5 Flash model.
• Math Context Normalization: Automatically detects spoken mathematical concepts and forcefully converts them into precise Unicode symbols.
• Local Data Management: An offline-first database offering custom labeling, multi-selection, pinning, duplication, and a secure trash recovery system.
• Custom Markdown Rendering: Built-in viewport capable of rendering headers, lists, typographical styles, and contextual text-string highlighting.
• Media Support: Functionality to import existing MP3/audio files for AI transcription, or export locally recorded sessions.
• Backup & Restore: Secure local database management via complete manual JSON export and restoration.

Installation Instructions:

1. Download the app-release.apk file attached below.
2. Install the APK on your Android device (you may need to allow installation from unknown sources).
3. Open the app, complete the onboarding process, and input your personal Google Gemini API Key to activate the AI processing features.