AI Agent Hub
Back to plugins
🧰

dsh-voice-input-cn

Web Tools Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install Schumchanvi/dsh-voice-input-cn

Paste the following prompt into your AI chat to install this plugin:

Run dsh plugin install Schumchanvi/dsh-voice-input-cn in the DeepSeek Harness terminal to install; source code is available at https://github.com/Schumchanvi/dsh-voice-input-cn .

About this plugin

The original DSH voice-input plugin depended on the Google Web Speech API, which is unreachable from mainland China. dsh-voice-input-cn swaps the recognition engine for Alibaba Cloud DashScope ASR and adds a small local Python bridge that handles SDK-level WebSocket authentication, making voice input genuinely usable behind the GFW.

The core experience: tap the microphone button beside the chat box and start talking. Recognized text is streamed into the textarea at the cursor position in real time without overwriting existing drafts. A 1.5-second silence window auto-stops recording. You can toggle between Chinese and English, and optionally enable auto-send after recognition completes. The bridge listens only on 127.0.0.1, audio is sent exclusively to DashScope for ASR, and nothing is stored or exposed publicly.

If you run DeepSeek Harness in mainland China and prefer dictating long prompts or messages over typing, this plugin closes that gap with a streaming recognition path that stays on a domestic, low-latency route to Alibaba Cloud.

Use Cases

  • Dictating long prompts in the DSH chat box without typing
  • Composing Chinese messages quickly with real-time cursor-position insertion
  • Bypassing unreachable Google Web Speech API by routing ASR through Alibaba DashScope

Best For

  • DeepSeek Harness users in mainland China
  • Developers who prefer dictating over extended typing
  • Users who need a domestic-cloud fallback when Google services are inaccessible