Releases: Makememo/MemoAI
Releases · Makememo/MemoAI
Memo AI v1.6.8
New Features
- Enhanced Search Experience: Select text and press Ctrl+F (Windows/Linux) or Cmd+F (macOS) to auto-fill the search box with selected content
- Video Playback Enhancement: Added flv.js support for playing FLV format video streams
- Homepage Improvements: Added view mode toggle, history filter, and optimized HistoryList component layout and performance
- Font System Upgrade: Implemented system font caching and validation, enhanced CJK font handling, added user notifications for font loading failures
- LLM Service Optimization: Updated model configurations and pricing for multiple LLM providers, enhanced parameter validation, improved model instantiation logic, implemented LLM service config cache clearing for immediate effect after updates
Improvements
- Transcription Post-processing: Implemented keyword replacement post-processing for accurate short keyword replacement
- Server Communication: Enhanced server polling mechanism with max retry limits, improved message sending safety across handlers
- Update Process: Optimized macOS update installation flow, added quit flag setting for better update experience
Fixes
- UI Adjustments: Optimized Fonts component layout and styling
Memo AI v1.6.7
Update README.md
Memo AI v1.6.6
Update README.md
Memo AI v1.6.3
Features
- Enhanced export project with an options parameter for more flexible exporting
- Updated yt-dlp to the latest version for better compatibility
Fixed
- Resolved a crash when exporting video
Memo AI v1.6.2
Features
- Whisper transcription upgrade: smarter punctuation/number merging, multi-language detection, full JSON & segments export.
- New LLM architecture completed with dynamic service/model loading and added providers (e.g. MistralAI, LM Studio, xAI).
- Chat enhancements: history management, adjustable parameters (temperature, max tokens), font size control, cancel (stop) support.
Improvements
- Better long audio stability and performance (GPU and buffer tuning).
- Refined UI/UX: cleaner layouts, improved language strings, smoother navigation.
Fixes
- Corrected export and transcription progress display.
- Resolved truncated output for long logs/content.
Memo AI v1.6.0
Memo AI v1.5.5
Features
- Support selecting the number of speakers during speaker recognition (local model needs to be re-downloaded)
- Text-to-speech synthesis speed increased by 5x
- Text-to-speech function upgraded with segmented algorithm
- Third-party integration supports uninstallation and reinstallation
Fix
- Fixed the issue with the O shortcut key not opening
- Fixed the error caused by the download source
- Fixed the issue where scrolling is not possible after replacing text with more than ten replacement words
- Fixed the issue where capitalization of the first letter of English changes to lowercase when replacing punctuation symbols
Memo AI v1.5.3
New Features
- Introduced background music separation to assist with transcription
- Added option to switch model download sources
- Implemented automatic retry for failed batch tasks
- Added re-export function for batch tasks
- Expanded subtitle optimization options
- Added subfolder creation feature (beta)
Optimizations
- Improved file reading and writing efficiency
- Enhanced batch file sorting mechanism
- Optimized TTS execution for batch tasks
- Improved stability of batch task processing
Memo AI v1.5.2
New Features
- Application: Accompaniment Extraction, supports separation of sound and music, please go to Settings-Lab Turn on
- Summary: Optimizes DeepSeek R1 thinking model output
- Transcription: Added Groq, enabling fast cloud-based voice-to-text conversion
Bug Fixes
- Fixed reading issues causing blank
Memo AI v1.5.1
Feature
- Supports m2t format
Optimize
- Fixed Open AI summary error 400
- Fixed Open AI summary button misalignment
- Fixed the issue where there is a space after downloading the link and it cannot be resolved

