# Chatterbox TTS

> Use this tool when you need to generate high-quality speech from text, such as providing spoken feedback or code reviews, with configurable expressiveness controls and flexible dual TTS engine support. It solves problems like automated voice interaction, daily engagement, and audio resource management, accepting text inputs and producing spoken audio outputs. Ideal for developers building AI assistants, it provides real-time progress updates and device-optimized performance.

Canonical page: https://skillsregistry.net/skills/digitarald-chatterbox-tts  
JSON: https://api.skillsregistry.net/v1/skills/digitarald-chatterbox-tts

## Description

This MCP server provides text-to-speech functionality using either the high-quality Chatterbox TTS neural model or macOS's built-in 'say' command, automatically generating speech from text and playing it back with configurable expressiveness controls. Built by digitarald using Python and FastMCP, it features automatic model loading with progress notifications, persistent audio file storage with configurable TTL cleanup, dual TTS engine support for flexibility between quality and speed, and specialized prompts for daily haikus and code roasts. The implementation handles device optimization (MPS/CUDA/CPU), provides real-time progress updates during speech generation, and includes comprehensive audio resource management with embedded playback, making it valuable for developers who want AI assistants to provide spoken feedback, code reviews, or daily engagement through voice interaction.

## Trust

- **Trust score (0–1):** 0.91
- **Verification tier:** verified
- **Last scanned:** 2026-09-28

## Facts

- **Version:** 1.0.0
- **Skill type:** atomic
- **Execution layer:** mcp-remote
- **Runtime environment:** api
- **Category:** media
- **Updated:** 2026-09-28

## Source

- **Source listing:** [PulseMCP](https://www.pulsemcp.com/servers/digitarald-chatterbox-tts)
- **Repository:** <https://github.com/digitarald/chatterbox-mcp>

## Use it

Resolve this record through the SkillsRegistry MCP server (no auth, read-only):

```
claude mcp add --transport http --scope user skillsregistry https://api.skillsregistry.net/mcp
```

```json
{
  "jsonrpc": "2.0",
  "id": 1,
  "method": "tools/call",
  "params": {
    "name": "get_skill",
    "arguments": {
      "slug": "digitarald-chatterbox-tts"
    }
  }
}
```

REST: `GET https://api.skillsregistry.net/v1/skills/digitarald-chatterbox-tts` · pull for local use: `GET https://api.skillsregistry.net/v1/skills/digitarald-chatterbox-tts/pull`

---
SkillsRegistry indexes agent skills from public registries and GitHub. Skills we have analysed are scanned with Circle-IR and scored on six dimensions; each listing states its scan coverage. More: https://skillsregistry.net/llms.txt