AI Agent Hub
Back to plugins
🖥️

dsh-voice-input

Client Updated 2026.08.25

Run the following command in DeepSeek Harness:

dsh plugin install NewDaNew/dsh-voice-input

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

Run dsh plugin install NewDaNew/dsh-voice-input in your terminal; the plugin source is at https://github.com/NewDaNew/dsh-voice-input . After installation, restart dsh web and refresh your browser, and a microphone button will appear beside the input field.

About this plugin

The DSH web input box is keyboard-only out of the box. dsh-voice-input adds a microphone button right beside it, so you can dictate your prompt and the transcribed text lands in the input field without a single keystroke.

All speech recognition runs locally in the browser via the Web Speech API. No audio leaves your machine, no API key is required, and no server is involved. You can pick the recognition language (Auto, Chinese, or English) and it defaults to your system locale. A live transcript appears above the button while you speak, and a second tap commits the text. Enable auto-send after recognition and the final click disappears entirely.

It targets anyone on DSH 0.1.0-rc.6 or later with Chrome or Edge. The plugin is especially handy for long prompts, rapid follow-up questions, or anyone who finds speaking faster than typing. The button automatically hides in hero state and disables during message submission, so it stays out of the way of your existing workflow.

Use Cases

  • Dictating long prompts in DSH web without typing
  • Submitting messages hands-free when typing is inconvenient
  • Toggling auto-send to go from speech to sent in one step

Best For

  • DSH web developers who prefer voice input
  • Prompt engineers who type long text frequently
  • Chrome or Edge users seeking a keyboard-free workflow