Générateur AI LRC

Changelog

AI LRC Generator version history and updates.

All notable changes to AI LRC Generator will be documented in this file.

[1.4.0] - 2026-09-07

Added

  • Direct Lyrics and Subtitle Export: AI recognition, manual editing, and batch tasks can now export LRC, SRT, ASS, WebVTT, and TTML directly without visiting a format converter page.
  • Word-Level Timing: AI analysis now offers a word-level mode that generates Enhanced LRC. The editor supports word highlighting and timing adjustments, while downloads can be converted to standard LRC for compatibility with regular players.
  • Multilingual Pronunciation Guides: Pronunciation guides can now be enabled independently of lyrics translation. Supported options include Japanese romaji and kana, Chinese Pinyin and Zhuyin, Cantonese Jyutping, Korean Revised Romanization, and IPA. Preview and export can combine original lyrics, pronunciation, and translation.
  • Reference Lyrics Review: Recognition results can now be compared line by line with pasted reference lyrics. Text and formatting differences are highlighted, and users can apply individual corrections or high-confidence changes in bulk while preserving timestamps and vocalist labels.

Optimizations and Enhancements

  • Export Content Options: Every lyrics and subtitle format can include or exclude translations, pronunciation guides, and chorus labels. Batch exports are packaged into a single ZIP file to reduce browser blocking of multiple downloads.
  • Translation and Pronunciation Quality: Stable lyrics line identifiers are now used to generate and validate translations and pronunciation guides, preventing reordered model responses from misaligning lyrics. Generation prompts and result validation have also been improved for Japanese and other languages.
  • Timestamp Quality: Improved timestamp offset handling for segmented long audio, added support for timestamps with a single-digit minute field, and strengthened validation for word timing order, audio duration boundaries, and lyrics line breaks.
  • Korean Lyrics Wrapping: On-site previews now prefer line breaks between Korean words instead of splitting a word. Matching constraints have also been added to AI recognition and subtitle export.
  • Reference Lyrics Matching: Improved confidence scoring and correction recommendations when matching reference lyrics against recognition results, making safe-to-apply differences easier to identify.

Fixes and Improvements

  • Download and Playback Compatibility: Fixed download links expiring too early in some browsers, batch downloads being blocked, and missing feedback when audio failed to load or play. The player now shows categorized errors and offers a retry action.
  • Expired Audio Recovery: Fixed historical audio remaining unplayable after it was uploaded again. Retrying now refreshes the audio access URL.
  • Lyrics Editing and Export: Fixed pronunciation markers being accidentally removed during editing, word-level timestamps being lost, and inconsistent output rules across subtitle converter entry points.

[1.3.0] - 2026-07-19

Added

  • AI Lyrics Generation: Added the AI Lyrics Generator, which creates lyrics from a creative prompt and provides a generation task history.
  • Lyrics Translation: Added a translation option to AI Lyrics Recognition, enabling bilingual LRC files with original and translated lines. Translation can also be configured for batch uploads.
  • Chorus and Vocal Part Labels: AI can now identify and label different vocalists and chorus sections automatically, with an option to quickly customize vocalist display names after recognition.
  • Video File Uploads: AI recognition and manual editing workflows now accept common video formats and can extract their audio tracks for lyrics processing.
  • Quick SRT Export: Lyrics can now be downloaded directly as SRT subtitles after recognition or editing. The LRC to SRT tool remains available for batch conversion.

Optimizations and Enhancements

  • LRC Download Options: Downloads can now retain or remove translations and chorus labels as needed for different players and use cases.
  • Open API: Expanded the Open API documentation with complete AI analysis parameters, including reference lyrics, language, lyrics translation, and chorus labeling. File size, request rate, and task queue limits have also been increased.
  • Workspace Sidebar: The task sidebar can now be collapsed to provide more space for lyrics editing.
  • AI Lyrics Generation Experience: Improved the generation page, task history, failure states, and limited-time free notices for clearer task feedback and a smoother creative workflow.
  • WeChat Pay: WeChat Pay now supports purchasing premium plans, with subscription plans provided on a non-renewing basis.
  • Feedback: Improved the timing and frequency of feedback prompts and added information about credit rewards for high-quality suggestions that are adopted.

Fixes and Improvements

  • Mobile Uploads: Fixed an issue in Chrome on iPhone where MP3 and WAV files appeared disabled and could not be selected.
  • Lyrics Processing: Fixed lyrics translation not taking effect, certain LRC parsing errors, and incomplete millisecond digits in timestamps.

[1.2.0] - 2026-05-16

Added

  • User API: Added User API interface for programmatic access and management.
  • Batch Processing: Improved batch processing workflow for better multi-file handling efficiency.
  • Credit System Upgrade: Added more credit pricing tiers and optimized credit quantity configuration. Visit the Pricing page for details.
  • Audio Storage Policy: Different user tiers enjoy different historical audio storage durations. Audio files are automatically cleaned up when expired, while processing records are retained.
  • AI Metadata Recognition: Backend optimization supports AI analysis of audio metadata, lyrics content, and language type.
  • Refund Policy: Added Refund Policy page to improve user rights explanation.
  • Documentation Center: Expanded Docs page content with more comprehensive usage guides.

Optimizations and Enhancements

  • AI Analysis Engine: Optimized the AI analysis engine performance, enhanced recognition capabilities, and significantly improved processing speed.
  • File Upload: Increased maximum file upload size limit, with gradual increases planned.
  • Queue Design: Optimized backend queue architecture to improve task scheduling efficiency.
  • Open API: Optimized Open API support and documentation.
  • Error Logging: Optimized error log output by removing redundant tags for better readability.

Fixes and Improvements

  • Magic Link: Fixed the issue where Auth.js magic link tokens were pre-fetched by email security scanners, causing verification failures.
  • Email Verification: Fixed email verification URL and verification page related issues; fixed duplicate email sending issue.
  • Feedback Popup: Fixed the issue where the feedback window reappeared after being closed.

[1.1.0] - 2026-03-02

Added

  • Classic Editor: Added a separate page display for the old LRC editor. Old LRC Editor

  • AI Recognition: The AI ​​recognition page is now independent, displayed using a workbench mode. AI Recognition

  • Manual Editing: The manual editing page is now independent, displayed using a workbench mode. Manual Editing

  • Credit History: Added a credit history page to conveniently view the history of credit changes. Credit History

Optimizations and Enhancements

  • User Interface: Deeply optimized the homepage UI and visual experience of all pages, and optimized the height of the left-hand list on mobile devices.

  • Internationalization: Completely revamped and improved multilingual content.

  • Task Panel: Optimized the task creation process and panel display information, supporting the display of audio metadata.

Fixes and Improvements

  • Display Fixes: Fixed the issue of reference lyrics not displaying in AI analysis tasks; fixed abnormal page number display.

  • Time Precision: Fixed the floating-point drift issue on the lyrics timeline.

[1.0.1] - 2025-08-18

Added

  • LRC Converter Toolkit: Added comprehensive LRC format conversion tools

    • LRC to SRT: Convert lyrics files to standard subtitle format for video production and podcast subtitles
    • LRC to ASS: Convert to advanced subtitle format with rich styling and effects support
    • LRC to WebVTT: Convert to HTML5 standard subtitle format, perfectly compatible with modern browsers
    • LRC to JSON: Convert to structured data format for programming development and API integration
    • LRC to TTML: Convert to W3C standard XML subtitle format for professional media production

  • SRT Converter Tools: Added SRT format conversion capabilities

    • SRT to LRC: Convert video subtitles to lyrics format, supporting karaoke apps and music players

Technical Features

  • Batch Processing: All conversion tools support batch file processing for significantly improved efficiency
  • Precise Timeline: Advanced algorithms ensure precise timeline accuracy during conversion
  • Format Compatibility: Support for multiple time formats and encoding methods ensuring broad compatibility
  • Real-time Preview: Live preview functionality for conversion results to verify quality

User Experience

  • Navigation Tools Grouping: Group conversion tools by format type for easier user navigation
  • Complete Documentation Support: Comprehensive documentation and help for all converter tool pages

[1.0.0] - 2025-07-13

Added

  • Core AI Lyrics Recognition: Implemented advanced AI-powered lyrics recognition system
  • Multi-language Support: Added support for English, Chinese, Korean, and Japanese lyrics
  • Real-time Editing Interface: Built intuitive interface for real-time lyrics editing with precise timing control
  • Batch Processing: Added capability to process multiple audio files simultaneously
  • LRC File Generation: Implemented automatic LRC file generation with precise timestamp synchronization
  • Audio Format Support: Support for MP3, WAV, FLAC, AAC, OGC audio formats
  • User Authentication: Implemented secure user authentication system
  • Export Options: Multiple export formats including standard LRC and custom formats
  • Responsive Design: Mobile-friendly interface design

Technical Features

  • AI Model Integration: Integrated state-of-the-art speech recognition and natural language processing models
  • Audio Processing Pipeline: Optimized audio processing for fast and accurate lyrics extraction
  • Time Synchronization: Millisecond-precision timestamp synchronization
  • Data Security: Implemented secure audio file processing with temporary storage

User Experience

  • Intuitive Interface: Clean and user-friendly design for easy navigation
  • Progress Indicators: Real-time progress tracking for audio processing
  • Preview Functionality: Live preview of lyrics with audio synchronization
  • Keyboard Shortcuts: Efficient keyboard shortcuts for power users
  • Help System: Comprehensive help documentation and tutorials

Security

  • Data Protection: Secure handling of user audio files
  • Privacy Compliance: GDPR-compliant data processing
  • Temporary Storage: Audio files processed in temporary environments
  • Access Control: Secure user authentication and authorization