Will Owl Meeting upload my recording data?

No. Owl Meeting adopts a 100% offline architecture; all speech recognition and AI processing are done locally, and data is stored in the user's local directory.

How do I back up or migrate my recognition models?

You can copy the files in the %LOCALAPPDATA%/owl_meeting/models directory to a new path and re-specify the model folder in the settings.

Back Home

🛡️ Data Storage & Privacy

Last Updated: 2026-04-21 · Language: English

1. Privacy Architecture Overview

Offline recognition engine: Speech recognition models run locally without internet access.
Offline AI processing: If you use locally deployed Ollama models, translation, correction, and summary tasks stay local.
No telemetry tracking: Owl Meeting does not upload usage logs by default.

2. Data Storage Locations

App Data: %LOCALAPPDATA%/owl_meeting (Includes database, search index, models, etc.).
Audio Outputs: Music/owl_meeting/audio.
Speaker Samples: Music/owl_meeting/speaker_audio.

3. Offline / Intranet Usage

Download the necessary recognition models in advance.
Download the necessary LLM models for AI features.

4. Model File Migration

Copy all files from %LOCALAPPDATA%/owl_meeting/models to a new folder of your choice.
Specify the new folder path in [Settings] -> [Models].
Restart the application.
Once verified, you may clear the contents of the original model directory.

5. Important Notes

Uninstalling the software will clear all application data (excluding individual audio export files).