home / skills / openclaw / skills / text-to-speech

text-to-speech skill

/skills/okaris/text-to-speech

This skill converts text to natural speech using inference.sh across multiple TTS models for voiceovers, podcasts, and accessibility.

npx playbooks add skill openclaw/skills --skill text-to-speech

Review the files below or copy the command above to add this skill to your agents.

Files (2)
SKILL.md
4.1 KB
---
name: text-to-speech
description: "Convert text to natural speech with DIA TTS, Kokoro, Chatterbox, and more via inference.sh CLI. Models: DIA TTS (conversational), Kokoro TTS, Chatterbox, Higgs Audio, VibeVoice (podcasts). Capabilities: text-to-speech, voice cloning, multi-speaker dialogue, podcast generation, expressive speech. Use for: voiceovers, audiobooks, podcasts, accessibility, video narration, IVR, voice assistants. Triggers: text to speech, tts, voice generation, ai voice, speech synthesis, voice over, generate speech, ai narrator, voice cloning, text to audio, elevenlabs alternative, voice ai, ai voiceover, speech generator, natural voice"
allowed-tools: Bash(infsh *)
---

# Text-to-Speech

Convert text to natural speech via [inference.sh](https://inference.sh) CLI.

![Text-to-Speech](https://cloud.inference.sh/u/4mg21r6ta37mpaz6ktzwtt8krr/01jz00krptarq4bwm89g539aea.png)

## Quick Start

```bash
# Install CLI
curl -fsSL https://cli.inference.sh | sh && infsh login

# Generate speech
infsh app run infsh/kokoro-tts --input '{"text": "Hello, welcome to our product demo."}'
```

> **Install note:** The [install script](https://cli.inference.sh) only detects your OS/architecture, downloads the matching binary from `dist.inference.sh`, and verifies its SHA-256 checksum. No elevated permissions or background processes. [Manual install & verification](https://dist.inference.sh/cli/checksums.txt) available.

## Available Models

| Model | App ID | Best For |
|-------|--------|----------|
| DIA TTS | `infsh/dia-tts` | Conversational, expressive |
| Kokoro TTS | `infsh/kokoro-tts` | Fast, natural |
| Chatterbox | `infsh/chatterbox` | General purpose |
| Higgs Audio | `infsh/higgs-audio` | Emotional control |
| VibeVoice | `infsh/vibevoice` | Podcasts, long-form |

## Browse All Audio Apps

```bash
infsh app list --category audio
```

## Examples

### Basic Text-to-Speech

```bash
infsh app run infsh/kokoro-tts --input '{"text": "Welcome to our tutorial."}'
```

### Conversational TTS with DIA

```bash
infsh app sample infsh/dia-tts --save input.json

# Edit input.json:
# {
#   "text": "Hey! How are you doing today? I'm really excited to share this with you.",
#   "voice": "conversational"
# }

infsh app run infsh/dia-tts --input input.json
```

### Long-form Audio (Podcasts)

```bash
infsh app sample infsh/vibevoice --save input.json

# Edit input.json with your podcast script
infsh app run infsh/vibevoice --input input.json
```

### Expressive Speech with Higgs

```bash
infsh app sample infsh/higgs-audio --save input.json

# {
#   "text": "This is absolutely incredible!",
#   "emotion": "excited"
# }

infsh app run infsh/higgs-audio --input input.json
```

## Use Cases

- **Voiceovers**: Product demos, explainer videos
- **Audiobooks**: Convert text to spoken word
- **Podcasts**: Generate podcast episodes
- **Accessibility**: Make content accessible
- **IVR**: Phone system voice prompts
- **Video Narration**: Add narration to videos

## Combine with Video

Generate speech, then create a talking head video:

```bash
# 1. Generate speech
infsh app run infsh/kokoro-tts --input '{"text": "Your script here"}' > speech.json

# 2. Use the audio URL with OmniHuman for avatar video
infsh app run bytedance/omnihuman-1-5 --input '{
  "image_url": "https://portrait.jpg",
  "audio_url": "<audio-url-from-step-1>"
}'
```

## Related Skills

```bash
# Full platform skill (all 150+ apps)
npx skills add inference-sh/skills@inference-sh

# AI avatars (combine TTS with talking heads)
npx skills add inference-sh/skills@ai-avatar-video

# AI music generation
npx skills add inference-sh/skills@ai-music-generation

# Speech-to-text (transcription)
npx skills add inference-sh/skills@speech-to-text

# Video generation
npx skills add inference-sh/skills@ai-video-generation
```

Browse all apps: `infsh app list`

## Documentation

- [Running Apps](https://inference.sh/docs/apps/running) - How to run apps via CLI
- [Audio Transcription Example](https://inference.sh/docs/examples/audio-transcription) - Audio processing workflows
- [Apps Overview](https://inference.sh/docs/apps/overview) - Understanding the app ecosystem

Overview

This skill converts text into natural, expressive speech using multiple TTS models (DIA TTS, Kokoro, Chatterbox, Higgs Audio, VibeVoice) via the inference.sh CLI. It supports voice cloning, multi-speaker dialogue, and long-form podcast generation. Use it to produce voiceovers, audiobooks, accessibility audio, IVR prompts, and podcast episodes. The CLI-driven workflow makes it easy to script, batch, and integrate into media pipelines.

How this skill works

The skill runs inference.sh apps for different TTS models by supplying JSON inputs that include text, voice selection, and optional parameters like emotion or speaker. You can sample model inputs, edit saved JSON, and run the chosen app to return audio files or URLs. Output audio can be chained into other apps (for example, a talking-head video generator) or downloaded for post-production. Voice cloning and multi-speaker dialogue are supported by providing speaker metadata and cloned voice samples where the model allows it.

When to use it

  • Create voiceovers for product demos, tutorials, and explainer videos
  • Produce audiobooks or narrated long-form content and podcasts
  • Generate accessible audio for websites and apps (screen readers, articles)
  • Build IVR prompts and voice assistants with consistent voice profiles
  • Scripted multi-speaker dialogue for simulations, demos, or training content

Best practices

  • Choose the model by intent: DIA TTS for conversational tone, VibeVoice for podcasts, Kokoro for fast natural reads
  • Start with model sample JSON, tweak parameters like emotion, pacing, and voice, then re-run for iterations
  • Use high-quality cloned voice samples and label speaker metadata for reliable multi-speaker outputs
  • Batch small segments for long-form audio to control pacing and make editing easier
  • Chain outputs into downstream apps by saving audio URLs rather than re-uploading large files

Example use cases

  • Turn product text and release notes into short narrated videos or social content
  • Convert novel chapters into audiobook files with consistent pacing and voices
  • Generate full podcast episodes from scripts, including intros, transitions, and multi-guest dialogue
  • Create IVR menus and phone prompts with clear, branded voice prompts
  • Produce accessible article audio for websites and e-learning modules

FAQ

Which model should I pick for a natural, conversational read?

Use DIA TTS for conversational and expressive delivery; Kokoro is a good default for fast, natural speech.

Can I clone a specific voice?

Yes—several apps support voice cloning. Provide the required sample audio and follow the model’s sample JSON for speaker configuration.

How do I generate long-form audio like podcasts without quality loss?

Split scripts into segments, generate each segment, and then concatenate in post. Use VibeVoice for podcast-optimized output and control pacing/emotion per segment.