dsh-funasr-voice
A DeepSeek Harness Web GUI plugin for fully local, offline microphone dictation using FunASR and SenseVoiceSmall.
Install
$ dsh plugin --profile <你的profile> add <本插件目录或仓库地址>Plugin Overview & Capabilities
AI-assisted organization based on the public repository snapshot. The content must be grounded in source evidence and does not replace compatibility or security verification.
Key Capabilities
- Captures microphone audio through AudioWorklet.
- Uses VAD to segment continuous dictation.
- Runs local FunASR with the SenseVoiceSmall model for offline speech recognition.
- Fills recognized text into the DSH input draft and can optionally send it automatically.
- Provides GUI settings and YAML configuration for Python and model paths.
- Can auto-detect configured local Python and SenseVoiceSmall locations.
- Removes SenseVoice event and emotion emoji from recognized text.
Useful For
- Offline voice dictation in the DeepSeek Harness Desktop Web GUI.
- Hands-free or faster drafting of DSH prompts.
- Local speech-to-text workflows where audio is not uploaded to a remote ASR service.
Who It Fits
- DeepSeek Harness Desktop users.
- Users who have a local Python, FunASR, and SenseVoiceSmall environment.
- Users seeking local Chinese and multilingual voice input.
Documented Limitations
- Requires DeepSeek Harness Desktop, with version 2.0.2 identified as the development/test version.
- Requires Node.js 20 or later on the plugin host side.
- Requires a separately installed Python virtual environment, FunASR 1.4.x, torch, and a local SenseVoiceSmall model.
- The README states that DSH Desktop 2.0.2 has a microphone-permission bug that can yield silent audio unless DSH Desktop is patched or officially fixed.
- Initial recognition may take about 2–3 seconds while the model and torch initialize.
- Recognition runs on local CPU according to the README.
DSH Compatibility
Version-specific runtime evidence collected by DSH Plugin. A missing result means we have not tested that combination yet.
Security Signals
Objective signals discovered from package metadata and source inspection. These are not a guarantee that a plugin is safe.
package.json declares 0 runtime, 0 development, 1 peer, and 0 optional dependencies.
package.json declares DSH bundle metadata.
GitHub reports the repository license as MIT.
A root package.json was captured and can be inspected by the registry.
Public GitHub source metadata is available for this registry snapshot.
Source & Registry Notes
Public provenance, Registry classification, and the latest source check for this entry, kept separate from runtime verification.
- Source repository
- fenglin-ai/dsh-funasr-voice
- Registry source
- GitHub · dsh-plugin topic
- Registry classification
- Plugin
- Source checked
- 8128391 · 2026-08-27
This project is independently indexed from public source information. DSH Plugin is not affiliated with DeepSeek or the plugin author. Always check the author repository before installation.
Repository Activity
- GitHub stars
- 1GitHub stars
- Forks
- 0
- Open issues
- 0
- Last commit
- 2026-08-26
- Last release
- No release detected
Related DSH Plugins
Ranked by overlapping capabilities, use cases, plugin type, categories, and DSH profile.
In-app visual plugin marketplace for DeepSeek Harness to discover, install, update, and manage community plugins.
View plugin →A customizable glassmorphism theme plugin for the DeepSeek Harness web UI.
View plugin →A terminal workspace and Claude Code style TUI surface for DeepSeek Harness.
View plugin →DeepSeek Harness web-search plugin that adds web search, X search, and focused page fetching with structured evidence.
View plugin →