Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

voicebox

voicebox• Pinned

voicebox: A local-first AI voice studio for cloning voices, generating speech, dictating into any app, and giving agents a voice you own.

What problem it solves

Voice input and voice output are usually handled by separate cloud services, which means audio and transcripts leave the machine and the two halves of the voice loop do not connect. Voicebox aims to cover both directions locally: cloning voices from short samples, generating speech, dictating into any text field, and letting MCP-aware agents speak back in a cloned voice.

Key features

  • Multi-engine voice cloning and speech generation — Eight switchable TTS engines cover 23 languages, with zero-shot cloning from a reference sample plus preset voices from Kokoro and Qwen CustomVoice. Qwen VoiceDesign can build a voice from a written description.

    Source: README

  • Global dictation and voice input — A configurable hotkey records from anywhere on the system and pastes the transcript into the focused text field, with in-app mic buttons on every Voicebox text field and Whisper-based transcription.

    Source: README, backend

  • Agent voice output over MCP and REST — A built-in Model Context Protocol server and REST endpoints let agents speak, transcribe, and browse captures and profiles, with per-client voice bindings and an always-visible on-screen pill.

    Source: README

  • Post-processing effects and expressive delivery — Eight audio effects powered by Spotify's pedalboard library can be previewed, chained, and saved as presets, while Chatterbox Turbo interprets paralinguistic tags such as [laugh] and [sigh].

    Source: README

  • Long-form generation, versioning and stories editor — Text is auto-chunked at sentence boundaries and crossfaded for scripts and chapters, each generation keeps tracked versions, and a multi-track timeline editor supports conversations, podcasts, and narratives.

    Source: README

Quick start

Copied verbatim from the project's README.

claude mcp add voicebox \
  --transport http \
  --url http://127.0.0.1:17493/mcp \
  --header "X-Voicebox-Client-Id: claude-code"

Who it may suit

This may suit developers and creators who want voice cloning, dictation, and agent speech to run on their own hardware rather than through a hosted service. It may also suit accessibility users who rely on speech input or need voices in place of their own.

Topics to discuss

  • Which additional TTS or STT engines should be added next, and what would make engine selection easier for non-experts?

  • How should the local LLM step be tuned so dictation cleanup and voice personalities do not distort technical or multilingual transcripts?

  • What should the plugin and pipeline-routing roadmap look like so third parties can add models, transforms, and sinks?

Written by ReadmeX AI from the project's README, source and public GitHub data; it may contain mistakes.

Read the original README https://github.com/jamiepine/voicebox#readme

ReadmeX opened this community for a trending project; it is not run by the project. Maintainers can claim it from the community page.

github.comhttps://github.com/jamiepine/voicebox

Comments

…