Local Vosk Speech-to-Text for Openclaw

Local Vosk STT provides a lightweight and fully offline speech-to-text solution for transcribing audio files and voice messages without relying on cloud-based APIs.

sfkiwi
v1.0.1
Feb 10, 2026
0
2.2k
0

Install & Download

1. ClawHub CLI

The fastest way to install a skill directly from the registry.

npx clawhub@latest install local-vosk

2. Manual Installation

Copy the skill folder to one of these locations

Global
~/.openclaw/skills/
Workspace
<project>/skills/

Priority: Workspace > Local > Bundled

3. Prompt Installation

Copy this prompt to OpenClaw to install it automatically.

Help me install local-vosk using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).

Prefer to download?

Get the raw skill files in a ZIP archive.

What is Local Vosk Speech-to-Text?

The Local Vosk STT skill brings high-performance, private transcription capabilities to your local environment. Built on the Vosk framework, it allows users to process audio files, including Telegram voice notes, in real-time or faster. By leveraging Openclaw Skills, developers can integrate voice processing into their workflows while maintaining 100% data privacy and zero API costs.

This skill is designed for those who require efficiency and reliability. Unlike cloud-based solutions, this tool remains functional without an internet connection, ensuring that sensitive audio data never leaves your local machine. It serves as a foundational component for building privacy-first AI assistants and automation tools.

Local Vosk Speech-to-Text Use Cases

  • Automatically transcribing Telegram voice messages in .ogg format.
  • Processing various audio formats like mp3, wav, and flac using ffmpeg.
  • Implementing speech-to-text in environments with restricted or no internet access.
  • Reducing latency and costs by avoiding third-party cloud transcription services.

How Local Vosk Speech-to-Text Works

  1. The user provides an audio file path to the transcription script via the command line.
  2. The system utilizes ffmpeg to decode the audio into a format compatible with the Vosk engine.
  3. The Vosk engine loads the pre-downloaded language model to analyze the audio stream.
  4. The skill performs the transcription locally, outputting the text content directly to the console or integrated workflow.

Local Vosk Speech-to-Text Setup

To get started with this skill, install the Vosk library and download a language model:

pip3 install vosk --user --break-system-packages

# Download and setup the default small model
mkdir -p ~/vosk-models && cd ~/vosk-models
wget https://alphacephei.com/vosk/models/vosk-model-small-en-us-0.15.zip
unzip vosk-model-small-en-us-0.15.zip

Local Vosk Speech-to-Text Data Schema & Taxonomy

Data Component Description
Local Models Stored in ~/vosk-models/; identifies linguistic patterns for recognition.
Input Support Processes any audio container supported by ffmpeg (ogg, mp3, wav, etc.).
Execution Scripts Transcription logic housed in ./skills/local-vosk/scripts/transcribe.

Local Vosk Speech-to-Text Advanced Features

  • Support for multiple languages by downloading additional Vosk models for specific locales.
  • High-speed processing reaching approximately 10x real-time on standard consumer hardware.
  • Direct compatibility with Telegram's .ogg voice message format for seamless bot integrations.
  • Minimal resource footprint (~40MB for the small model) ideal for edge deployments.

SKILL.md


Loading

Related Openclaw Skills

METADATA

Github Stars: 0
forks: 0

Featured*