voicebox
tivojn/voicebox · skills.sh
Open source Repository Open in the app JSON README (API)
About
Skill publicada por tivojn/voicebox no skills.sh. Instale com: npx skills add tivojn/voicebox@voicebox
Details
- Kind
- Agent skills
- Publisher
- tivojn
- Origin
- skillssh
- Category
- ferramentas
- Last push
- 2026-03-23T00:16:14Z
- Repository state
- ativo
- Language
- Python
- Added
- 2026-10-07 06:31:25
- Updated
- 2026-10-07 06:31:25
- Origin id
tivojn/voicebox/voicebox
README
# Voicebox All-in-one voice toolkit for Claude Code on Apple Silicon Macs. Apple Silicon Mac 上的 Claude Code 一站式语音工具包。 ## Features / 功能 - **Voice Design / 语音设计** — Create custom voices from text descriptions / 用文字描述创建自定义声音 - **Voice Cloning / 语音克隆** — Clone any voice from a 10s audio sample / 录 10 秒音频克隆任意声音 - **Text-to-Speech / 文字转语音** — Generate speech in 10 languages / 支持 10 种语言的语音合成 - **Multi-Speaker Conversations / 多角色对话** — Generate dialogues, dramas, news broadcasts / 生成对话、戏剧、新闻播报 - **Transcription / 语音转文字** — Speech-to-text for audio & video, 52 languages with auto-detection / 音频视频转文字，52 种语言自动识别 ## Install / 安装 ```bash git clone https://github.com/tivojn/voicebox.git ~/.claude/skills/voicebox ``` That's it. Dependencies and models auto-download on first use. 就这样。依赖和模型首次使用时自动下载。 ## Requirements / 环境要求 - Apple Silicon Mac (M1/M2/M3/M4) - [Claude Code](https://claude.com/claude-code) - [uv](https://docs.astral.sh/uv/) (usually pre-installed with Claude Code / 通常随 Claude Code 预装) - ffmpeg — for recording & video transcription / 录音和视频转写需要 (`brew install ffmpeg`) ## Quick Start / 快速开始 ``` /voicebox create a calm narrator voice profile # Design a voice from description /voicebox "Calm Narrator" "Hello, this is a test." # Generate speech /voicebox clone my voice # Record from mic & clone /voicebox clone my voice from /path/to/audio.wav # Clone from audio file /voicebox transcribe /path/to/audio.wav # Transcribe audio/video /voicebox create a news broadcast with anchor and reporter # Multi-speaker conversation ``` ## Voice Profiles / 语音档案 Profiles are stored in `data/profiles.json`. Two types: - **Designed** — Created from a text description (supports `--instruct` style overrides) - **Cloned** — Created from a real audio sample (reproduces the tone/energy of the original recording) Start with no profiles — create your own with `/voicebox create ...` or `/voicebox clone ...`. ## Quality Options / 质量选项 All commands default to **high** (1.7B models). Use `--quality standard` for faster 0.6B models if needed. | Category / 类别 | High (default) | Standard | |-----------------|---------------|----------| | Voice Design / 语音设计 | 1.7B | (same) | | Voice Clone / 语音克隆 | 1.7B (~3.5GB) | 0.6B (~1.5GB) | | Transcription / 语音转文字 | 1.7B (~3.5GB) | 0.6B (~1.5GB) | ## Recording / 录音 When recording from the mic, the script: - Records for the specified duration (default 10 seconds) - Prints `DONE — recording complete!` and plays a system bell when finished - Auto-transcribes the recording using Qwen3-ASR - Creates a cloned voice profile automatically ## Supported Languages / 支持语言 **TTS:** English, Chinese, Japanese, Korean, German, French, Russian, Portuguese, Spanish, Italian **ASR:** 52 languages with auto-detection / 52 种语言自动识别 ## Script Commands / 命令参考 ```bash # List profiles uv run ~/.claude/skills/voicebox/scripts/voicebox.py list # Generate speech uv run ~/.claude/skills/voicebox/scripts/voicebox.py generate "Profile" "text" --play # Generate with standard quality (faster, less RAM) uv run ~/.claude/skills/voicebox/scripts/voicebox.py generate "Profile" "text" --quality standard --play # Create designed voice uv run ~/.claude/skills/voicebox/scripts/voicebox.py create-designed "Name" --desc "description" --lang en # Clone from audio file uv run ~/.claude/skills/voicebox/scripts/voicebox.py create-cloned "Name" --audio /path/to.wav --ref-text "transcript" --lang en # Record from mic and clone uv run ~/.claude/skills/voicebox/scripts/voicebox.py record "Name" --duration 10 --lang en # Transcribe audio/video uv run ~/.claude/skills/voicebox/scripts/transcribe.py /path/to/file.wav # Multi-speaker conversation from JSON script uv run ~/.claude/skills/voicebox/scripts/voicebox.py conversation /tmp/script.json --play # Delete a profile uv run ~/.claude/skills/voicebox/scripts/voicebox.py delete "Name" ```