Whisper Model Download, Installation & Selection Guide

Why Install Whisper Models?

V2sub performs speech recognition 100% locally on your computer. Because neural ASR models range from 500 MB to 3 GB, they are downloaded separately from the core installer. You only need to install a model once—future software updates will reuse your installed models.

---

Method 1: One-Click In-App Download (Recommended)

1. Open V2sub 2.0 and navigate to Config → Model Config.
2. In the Model Management table, click Download next to the model you want (small, medium, large, or large-turbo).
3. New in v2.0.2: Once a download finishes, the model dropdown in both Task Queue and Real-Time Player refreshes immediately without restarting the app!

---

Method 2: Manual Offline Installation from Cloud Drive

1. Download the model archive (.zip) from the 模型下载 (Models) folder in our Cloud Drives.
2. Extract the archive so you have a single folder (e.g., small, medium, large) containing model.bin directly inside it.
3. In V2sub Config → Model Config, click Open Model Directory and copy the extracted folder into the models directory.