About ditto.audio
ditto.audio is a text-to-speech MCP server for agents and other MCP clients. It lets them browse available voices and languages, synthesize audio, and receive temporary download links rather than audio bytes in chat.
Key features
- 01Browse available and owned voices
- 02List language and engine support
- 03Synthesize speech with download links
- 04Queue batches of up to 1,000 texts
- 05Check, retrieve, or cancel batches
Use cases
- 01Generating speech from text for download
- 02Choosing a voice and supported language
- 03Submitting many text-to-speech jobs together
MCP support
Exposes voice and language browsing, speech synthesis, and batch-management tools through a Streamable HTTP MCP endpoint. Users authenticate with a ditto.audio account or, for headless use, an API key.
Configuration
Remote servers need no local install. Pick your client, then copy the config or install in one click.
https://mcp.ditto.audioclaude mcp add --transport http audio-ditto-tts https://mcp.ditto.audio{"mcpServers":{"audio-ditto-tts":{"type":"http","url":"https://mcp.ditto.audio"}}}Open in Cursor{"servers":{"audio-ditto-tts":{"type":"http","url":"https://mcp.ditto.audio"}}}Open in VS Code{"mcpServers":{"audio-ditto-tts":{"type":"http","url":"https://mcp.ditto.audio"}}}{"mcpServers":{"audio-ditto-tts":{"serverUrl":"https://mcp.ditto.audio"}}}Run commands in a terminal; paste JSON into the client’s MCP config and restart. Servers that need sign-in will prompt on first connect.
https://mcp.ditto.audioAuth requiredTools
Details
From the project READMEditto.audio MCP server
Text-to-speech for AI agents. Connect ditto.audio to Claude, ChatGPT, Cursor, VS Code, Codex or any Model Context Protocol client. Your agent can then browse voices and synthesize speech that comes back as short-lived download links, never as audio bytes in the chat.
Endpoint: https://mcp.ditto.audio (Streamable HTTP)
You sign in with your ditto.audio account the first time your client connects. There is no key to paste. Each connection appears under Account -> API keys as a connected app, draws on the same monthly allowance as your keys, and can be disconnected there.
Install
| Client | How |
|---|---|
| Claude (web, desktop, mobile, and Claude Code signed in with a Claude plan) | Add to Claude, then Connect and sign in |
| Cursor | Add to Cursor |
| VS Code | Install in VS Code |
| Claude Code (API key or Console login) | claude mcp add --transport http --scope user ditto https://mcp.ditto.audio, then /mcp -> Authenticate |
| Claude Code (plugin) | /plugin marketplace add ditto-audio/ditto-mcp then /plugin install ditto@ditto-audio |
| ChatGPT | Settings -> Security and login -> Developer mode on, then chatgpt.com/plugins -> + -> https://mcp.ditto.audio |
| Codex | codex mcp add ditto --url https://mcp.ditto.audio then codex mcp login ditto |
Headless (CI, servers): skip the sign-in and send an API key from
your API keys page as Authorization: Bearer ditto_sk_live_....
Tools
| Tool | What it does |
|---|---|
list_voices | Browse the voice catalogue, and your own cloned or trained voices (marked owned) |
list_languages | Every language code and which engines speak it |
text_to_speech | Synthesize text and get a download URL (MP3 by default; FLAC/WAV and more on request) |
create_speech_batch | Queue up to 1000 texts at once |
get_speech_batch | Progress of a batch |
get_speech_batch_results | Download URLs for a batch's items |
cancel_speech_batch | Stop a batch; finished items keep their audio |
Full reference, generated from the live server: https://ditto.audio/docs/mcp
What is in this repository
| Path | For |
|---|---|
server.json | The official MCP Registry entry (audio.ditto/tts) |
.claude-plugin/marketplace.json, plugins/ditto/ | Claude Code plugin marketplace |
.cursor-plugin/plugin.json, mcp.json | Cursor Marketplace plugin |
Support
Email [email protected] or see ditto.audio/support. Using the server is subject to the terms and the privacy policy.
Copyright (c) 2026 ditto.audio. All rights reserved (see LICENSE).
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.