Owl Meeting

Many people have used speech-to-text features from Doubao or Qianwen, but some cannot use online services for such needs due to company regulations and other restrictions. Owl Meeting is a speech-to-text app where all functions including recognition and storage run locally.

Key Features:

1. Supports Mandarin, Chinese dialects, English, Japanese, Korean and 25 European languages.
2. Transcribes while recording, and supports simultaneous recording of microphone and system audio.
3. Recognizes all types of audio/video files. Built-in tools handle channel adjustment, audio extraction and other related issues.
4. No GPU required. In CPU mode, processing 30 minutes of audio takes only 1 minute (tested on i5-11400H).
5. Speaker diarization. Supports assigning different recognition models to each speaker in the same audio to improve accuracy.
6. Powerful text editing functions.
7. Custom dictionary. Automatically processes recognized text based on dictionary entries.
8. Supports batch deletion and replacement, with synchronized dictionary updates.
9. Click recognized text to auto-play the corresponding audio, so you can edit while listening.