Text to speech MCP server for AI tools

A remote MCP server that lets Claude, Cursor and other MCP clients speak text with Speakvora voices.

Tools

ToolWhat it does
speakTurns text into speech with a catalog voice and returns an audio URL. Numbers must be written out in words for languages other than English.
list_voicesLists the voices and the languages they speak.
search_phrasesSearches the checked phrase library by words, so stored (instant) phrases can be reused.

Connect it

Claude Code:

claude mcp add --transport http speakvora https://speakvora.com/api/mcp --header "x-api-key: YOUR_KEY"

Cursor and other clients that read mcp.json:

{
  "mcpServers": {
    "speakvora": {
      "url": "https://speakvora.com/api/mcp",
      "headers": {
        "x-api-key": "YOUR_KEY"
      }
    }
  }
}

Get a free key at the portal. Keep keys on a server; never put one in a public repository.

Questions

Is there an MCP server for text to speech?

Yes. Speakvora runs a remote MCP server (streamable HTTP) at https://speakvora.com/api/mcp. It has three tools: speak, list_voices and search_phrases. It needs a Speakvora API key sent as the x-api-key header.

How do I add it to Claude Code?

Run: claude mcp add --transport http speakvora https://speakvora.com/api/mcp --header "x-api-key: YOUR_KEY".

Where is it listed?

The official MCP registry as io.github.eyeguide-dev/speakvora, Smithery, and Glama (ownership verified). SDKs for Node (npm install speakvora) and Python (pip install speakvora) are on npm and PyPI.

What does it cost?

Catalog voices are $3 per million characters. Every account gets 100,000 free characters that play stored library phrases; new text needs a card. See the pricing page for the full list.

What are the limits?

Voices are AI-generated. Non-English clips are marked native_review_pending until a native speaker has reviewed them. The MCP tools use catalog voices; custom voices and agents are set up in the portal.

Get a free keyPricingDocs