# ditto.audio

> ditto.audio is a freemium MCP server for design and media. Text-to-speech for AI agents: browse voices and synthesize speech as download links. It runs as a hosted remote server, so there is nothing to install locally.

- **Kind:** MCP server
- **Pricing:** Freemium
- **Platforms:** API, Web
- **Website:** https://ditto.audio/docs/mcp
- **Repository:** https://github.com/ditto-audio/ditto-mcp
- **MCP Registry name:** audio.ditto/tts
- **Categories:** design and media
- **Listing updated:** 2026-11-15

## About

Audio never comes back as bytes in the chat. Each synthesis returns a short-lived download link, MP3 by default with FLAC, WAV and other formats on request, along with a JSON result that carries voice ids and URLs. Voices are referred to by id from the catalogue, for example kokoro-af-heart, not by display name, and the catalogue also lists your own cloned or trained voices marked as owned.

A batch mode queues up to 1,000 texts at once, with tools to check progress, fetch download URLs for each item and cancel a batch while keeping finished audio. A language listing shows which engines speak each language and how many native voices exist, so an agent can choose a voice before synthesizing.

Sign-in is through your ditto.audio account, so there is no key to paste. Each connection shows up under Account, API keys as a connected app, shares the monthly allowance of your keys and can be disconnected there. For CI and servers, an API key can be sent as a bearer token instead. The README states the code is all rights reserved.

## Features

- **Voice catalogue:** Browse voices, including your own cloned or trained ones.
- **Language lookup:** See which engines speak each language and how many native voices exist.
- **Link-based audio:** Speech arrives as short-lived download URLs rather than audio in the chat.
- **Batch narration:** Queue up to 1,000 texts and track or cancel the batch.
- **Several formats:** MP3 by default, with FLAC, WAV and more on request.
- **No key to paste:** Sign in with your account or use an API key for headless runs.

## Good for

- Narrate a script with a chosen voice
- Generate audio for many lines of text in one batch
- Pick a voice in a particular language and accent
- Produce voiceover files from an agent workflow
- Use a cloned voice you own for narration

## Install

### Claude Code (Terminal)

```sh
claude mcp add --transport http tts https://mcp.ditto.audio
```

### Claude Desktop (Settings → Connectors → Add custom connector)

```sh
https://mcp.ditto.audio
```

### Cursor (~/.cursor/mcp.json)

```json
{
  "mcpServers": {
    "tts": {
      "url": "https://mcp.ditto.audio"
    }
  }
}
```

### VS Code (.vscode/mcp.json)

```json
{
  "servers": {
    "tts": {
      "type": "http",
      "url": "https://mcp.ditto.audio"
    }
  }
}
```

### Windsurf (~/.codeium/windsurf/mcp_config.json)

```json
{
  "mcpServers": {
    "tts": {
      "serverUrl": "https://mcp.ditto.audio"
    }
  }
}
```

### Other clients (your client's MCP settings (most use this shape))

```json
{
  "mcpServers": {
    "tts": {
      "type": "http",
      "url": "https://mcp.ditto.audio"
    }
  }
}
```

## Tools

- `list_voices`: Browses the voice catalogue, including your own cloned or trained voices.
- `list_languages`: Lists every language code and which engines speak it.
- `text_to_speech`: Synthesizes text and returns a download URL.
- `create_speech_batch`: Queues up to 1,000 texts at once.
- `get_speech_batch`: Returns the progress of a batch.
- `get_speech_batch_results`: Returns download URLs for a batch's items.
- `cancel_speech_batch`: Stops a batch while finished items keep their audio.

## FAQ

### Do I need to paste an API key?

No for interactive clients; you sign in with a ditto.audio account. Servers and CI can send an API key as a bearer token.

### How is audio delivered?

As short-lived download links, never as audio bytes in the chat.

### How many texts per batch?

Up to 1,000 at once through create_speech_batch.

### Which clients are supported?

Claude, ChatGPT, Claude Code, Cursor, VS Code and Codex, plus any Streamable HTTP MCP client.

### Is ditto.audio free?

ditto.audio has a free plan, with paid plans for more usage or features.

### How do I add ditto.audio to Claude Code?

Run this command in your terminal: claude mcp add --transport http tts https://mcp.ditto.audio

### What are the best alternatives to ditto.audio?

The closest alternatives to ditto.audio are HyperFrames by HeyGen, ffmpeg-skill, AvatarSay — Fictional Character & Talking Video MCP, ChartAnt and VfxSmith.

## Alternatives

- [HyperFrames by HeyGen](https://ai-powered.com/mcp/hyperframes-by-heygen): HyperFrames by HeyGen is a freemium MCP server for design and media.
- [ffmpeg-skill](https://ai-powered.com/mcp/ffmpeg-skill): ffmpeg-skill is an open-source MCP server for design and media.
- [AvatarSay — Fictional Character & Talking Video MCP](https://ai-powered.com/mcp/avatarsay-fictional-character-and-talking-video-mcp): AvatarSay — Fictional Character & Talking Video MCP is a freemium MCP server for design and media.
- [ChartAnt](https://ai-powered.com/mcp/chartant): ChartAnt is a freemium MCP server for design and media.
- [VfxSmith](https://ai-powered.com/mcp/vfxsmith): VfxSmith is a freemium MCP server for design and media.
- [CalderFlow](https://ai-powered.com/mcp/calderflow): CalderFlow is a free MCP server for design and media.
- [Eikino](https://ai-powered.com/mcp/eikino): Eikino is a freemium MCP server for security.
- [Hourlore](https://ai-powered.com/mcp/hourlore): Hourlore is a freemium MCP server for design and media.

---
Source: https://ai-powered.com/mcp/ditto-audio
