# Sensevoice Stt

> Offline speech-to-text transcription using SenseVoice. Fast (~1s cold start), lightweight (287MB), supports Chinese, English, Japanese, Korean and Cantonese. Rust/ONNX powered alternative to Whisper. Use when this capability is needed.

- Skill: `tomevault-io/sensevoice-stt` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add tomevault-io/sensevoice-stt`
- Raw SKILL.md: https://api.skillmd.com/api/skills/tomevault-io/sensevoice-stt/raw
- Safety review: pending (external: skill-scanner PASS, skillspector PASS)
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Coding & Dev Tools
- Author: tomevault-io (https://skillmd.com/u/tomevault-io)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/tomevault-io/sensevoice-stt

---


# SenseVoice Speech-to-Text

离线语音转文字插件，基于阿里巴巴 SenseVoice 模型 + Rust/ONNX Runtime。

## 特点
- 冷启动 ~1 秒（Whisper 需 30-40 秒）
- 总体积 287MB（Whisper ~1.5GB）
- 无常驻内存占用
- 中文转写质量优化
- 原生粤语支持

## 使用
语音消息自动转写，或使用命令：
- `/sensevoice setup` — 下载模型并配置
- `/sensevoice status` — 检查安装状态
- `/sensevoice test <文件>` — 手动测试转写

---
> Source: [xxtluuu/openclaw-sensevoice](https://github.com/xxtluuu/openclaw-sensevoice) — distributed by [TomeVault](https://tomevault.io).
<!-- tomevault:4.0:skill_md:2026-05-23 -->