New Features
• Large File Chunking Support - Automatic splitting of large audio files for transcription endpoints with file size limits (like OpenAI's 25MB limit). This only applies if you are not using the ASR service and use an API endpoint with a filesize restriction. To enable and configure this feature, add the following variables to your .env file:
# --- Large File Chunking (for endpoints with file size limits) ---
# Enable or disable the feature. It's on by default.
ENABLE_CHUNKING=true
# Set the max file size in MB before chunking is triggered.
# Use a value lower than your API's limit (e.g., 20 for OpenAI's 25MB limit).
CHUNK_SIZE_MB=20
# Set a few seconds of overlap to avoid cutting off words between chunks.
CHUNK_OVERLAP_SECONDS=3• Optimized File Processing - Eliminated uncompressed WAV file creation, WebM recordings stay as WebM, conversions use 32kbps MP3 for optimal size/quality balance
Bug Fixes
• Enhanced CSRF Protection - Fixed session timeout issues and improved security for long recording sessions
• Fixed File Size Setting Adherence - Application now properly respects configured file size limits
• Improved Recording Reliability - Fixed errors related to in-app recording and long recording sessions
• Vue.js Console Errors Fixed - Resolved template property warnings for better development experience