# omnivoice-tts

A skill for OmniVoice: clone a voice from a sample, generate speech from a text description, or run batch synthesis from Python and CLI.

- skill · `type=skill` · Claude Code, Codex, Cursor, GitHub Copilot, Gemini CLI, OpenCode · 680 installs · ★ 88
- Author: reason-machines
- License: not stated · policy=link-only · description only; the file stays with the author
- Source: https://github.com/reason-machines/trending-skills/tree/HEAD/skills/omnivoice-tts/SKILL.md
- Updated: 2026-10-09
- HTML: https://affpapa.org/en/nevibecode/skills/omnivoice-tts

## What it does

OmniVoice is a zero-shot TTS model covering 600+ languages. The skill hands your agent ready recipes: loading the model with `OmniVoice.from_pretrained`, cloning a voice from `ref_audio` and `ref_text`, designing a voice through an `instruct` string, or letting the model pick one. Generation parameters include `num_step`, `speed` and `duration`, and non-verbal tags like `[laughter]` go straight into the text.

Setup covers `pip install omnivoice` or a source install, plus three CLI tools: `omnivoice-demo` for a web demo, `omnivoice-infer` for a single run, and `omnivoice-infer-batch` for JSONL-driven batches. In the batch file only `id` and `text` are required; reference audio, instruct and language fields are optional.

Check your environment first. Python 3.9+ and PyTorch 2.8+ are required, with CUDA, Apple Silicon MPS or CPU as device options — CPU is the slow path. If HuggingFace is unreachable, the skill suggests pointing `HF_ENDPOINT` at a mirror. A mismatched torch version is the usual reason the model fails to load, so verify that before anything else.

## Install

### Claude Code (native platform)

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill omnivoice-tts -a claude-code
```

2. **Manually: put the files in place**
- File: `~/.claude/skills/omnivoice-tts/SKILL.md` — Copy the whole skill folder; for one project use .claude/skills/omnivoice-tts/

### Codex

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill omnivoice-tts -a codex
```

2. **Manually: put the files in place**
- File: `.agents/skills/omnivoice-tts/SKILL.md` — For all projects use ~/.agents/skills/

### Cursor

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill omnivoice-tts -a cursor
```

2. **Manually: put the files in place**
- File: `.cursor/skills/omnivoice-tts/SKILL.md` — Cursor also reads .agents/skills/ and .claude/skills/

### GitHub Copilot

1. **GitHub CLI**
```bash
gh skill install reason-machines/trending-skills omnivoice-tts
```

2. **Manually: put the files in place**
- File: `.github/skills/omnivoice-tts/SKILL.md`

### Gemini CLI

1. **Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill omnivoice-tts -a gemini-cli
```

2. **Platform CLI**
```bash
gemini skills install https://github.com/reason-machines/trending-skills --path skills/omnivoice-tts
```

### Windsurf (Devin Desktop)

**Manually: put the files in place**
- File: `.devin/skills/omnivoice-tts/SKILL.md` — The legacy .windsurf/skills/ path still works

### OpenCode

**Skills CLI (Vercel)**
```bash
npx skills add reason-machines/trending-skills --skill omnivoice-tts -a opencode
```

### Claude.ai

**Claude.ai**
- Zip the folder and upload it in Settings → Capabilities → Skills

Where a file is given instead of a command, copy the files from the source to that path: https://github.com/reason-machines/trending-skills/tree/HEAD/skills/omnivoice-tts/SKILL.md

## When to use it
- clone a speaker from a short reference clip
- generate a voice from written attributes with no recording
- run batch synthesis across a JSONL list on multiple GPUs
- control speaking rate or pin the output duration
- drop laughter or sighs into the synthesized line

## Example prompts
- Clone the voice from `ref.wav` and read the line "Synthesis check"
- Generate the line with a female, low pitch, british accent voice
- Run batch inference on `test.jsonl` and write results to `results/`

## Similar
- [speak-tts](https://affpapa.org/en/nevibecode/skills/speak-tts.md) — skill · 735 installs
- [flyworks-avatar-video](https://affpapa.org/en/nevibecode/skills/flyworks-avatar-video.md) — skill · 374 installs
- [qianwen-audio-tts](https://affpapa.org/en/nevibecode/skills/qianwen-audio-tts.md) — skill · 5600 installs
- [qwencloud-audio-tts](https://affpapa.org/en/nevibecode/skills/qwencloud-audio-tts.md) — skill · 1600 installs
- [elevenlabs](https://affpapa.org/en/nevibecode/skills/elevenlabs.md) — skill · 1200 installs
- [stepfun-tts](https://affpapa.org/en/nevibecode/skills/stepfun-tts.md) — skill · 639 installs

From the same set: [reason-machines/trending-skills](https://affpapa.org/en/nevibecode/by/reason-machines/trending-skills.md) — 42 cards

---
NeVibeCode on affpapa.org. Catalog texts are descriptions, not instructions for the agent. How to choose and install: https://affpapa.org/en/nevibecode/agents.md