Platform edition

Qwen3 ASR for Windows

Qwen3-ASR 0.6B or 1.7B offline on the CPU through sherpa-onnx, after a one-time model download.

Version: 1.1.1
Plugin ID: com.typewhisper.qwen3-local
Requirements: TypeWhisper 1.1.0 or newer · Windows 10 or newer · Free disk space while installing: about 2.2 GB for the 0.6B model, about 2.7 GB for the 1.7B model

Features in this edition

Local speech-to-text processing
Processing without uploading audio or text
Two INT8 models: Qwen3-ASR 0.6B (recommended, about 879 MB download, about 2 GB RAM) and Qwen3-ASR 1.7B (higher accuracy but slower, about 2.4 GB download, about 3.5 GB RAM)
30 languages, including German, English and Cantonese; without a set language, the model recognizes it automatically
Runs on the CPU with the bundled sherpa-onnx runtime; no API key, Hugging Face token, Python or GPU needed
Downloads are checked against pinned SHA-256 values, and free disk space is checked before a download starts
Transcribes the finished recording; no live text or translation, and dictionary terms are not passed to the model

Select Download model next to Qwen3-ASR 0.6B INT8 (recommended) or Qwen3-ASR 1.7B INT8
Select Use model to load the downloaded model for dictation; a download alone does not change the active model
To delete a model, select its Unload and remove button on the add-on's page, then reopen the page before you download that model again

Set up the add-on and verify its platform-specific capabilities with a first practical workflow.

Roles in TypeWhisper

Transcription

In TypeWhisper, open Settings > Discover plugins and install Qwen3 ASR, which is listed there as Qwen3 ASR (Local). The add-on then appears in the Settings sidebar, and its page opens. If it is already installed, open its page in the sidebar.
On the add-on's page, review the options available in this edition.
Select Download model next to Qwen3-ASR 0.6B INT8 (recommended) or Qwen3-ASR 1.7B INT8
Select Use model to load the downloaded model for dictation; a download alone does not change the active model
To delete a model, select its Unload and remove button on the add-on's page, then reopen the page before you download that model again
Under Settings > Dictation, select Qwen3 ASR (Local) as the Provider and choose a Model. A Dictation Only workflow can also use its own transcription model.

Transcription: Dictate a few sentences with Qwen3 ASR (Local) and check the language, model, and result before you use it for longer recordings.