v3.3.0 2025
cheaper-inference/OmniRouteJan 12, 2026 → Feb 6, 202625 days between releases
What's New
Shift from form filling automation to speech processing with Docker and MPS support
Breaking Changes
- Refactored folder structure and argument handling which may break existing imports or CLI usage
- Major architectural refactor of handlers impacting core functionality
New Features
- Added Docker container support for easier deployment
- Added MPS (Metal) backend support for Apple Silicon
- Added Paraformer ZH ASR support for Chinese speech recognition
- Added ChatTTS support for text-to-speech generation
- Added DeepFilterNet for speech enhancement
Improvements
- Improved logging for better debugging
- Fixed missing sounddevice module dependency
- Fixed issue with NLTK library
- Minor documentation updates
Bug Fixes
- Fixed NLTK library issue
- Fixed missing sounddevice module
Upgrade Notes
The project has refactored its folder structure and argument handling, which may require updating import paths or CLI usage. Additionally, the focus has shifted to speech processing, so existing automation workflows may not be compatible.
The Two Releases
Release v3.3.0
Jan 12, 2026· 9 months ago
Focuses on improving automation reliability by refining form filling logic and key press handling to prevent misidentification and improve data entry.
Full release notes2025 release
Feb 6, 2026· 9 months ago
An early foundation release focusing on Dockerization, MPS support, and the integration of various speech backends like Paraformer and ChatTTS.
Full release notes