Gandr TTS
Official# gandr-mcp
The official **Gandr MCP server**, text to speech inside any MCP-compatible agent (Claude Desktop, Cursor, and friends).
[](https://smithery.ai/servers/gandr/gandr-mcp)
## Tools
| Tool | What it does |
|---|---|
| `synthesize` | Render text to speech (WAV, base64). Voice, language, sample rate 8000-48000, pitch and pacing dials. |
| `list_voices` | The six available voices. |
| `list_languages` | The 23 supported languages and their codes. |
| `get_usage` | Tokens used by the configured key. One token is one character. |
## Install
```bash
pip install gandr-mcp
```
Claude Desktop (add to `claude_desktop_config.json`):
```json
{
"mcpServers": {
"gandr": {
"command": "gandr-mcp",
"env": {
"GANDR_API_KEY": "gnd_..."
}
}
}
}
```
Get a key at [gandr.ai](https://gandr.ai). Docs: [gandr.ai/docs](https://gandr.ai/docs).
---
mcp-name: io.github.Gandr-AI/gandr-mcp
TDQS
Scored across 4 tools
Each tool has a clearly distinct purpose: synthesize for TTS, list_voices for voice metadata, list_languages for language metadata, and get_usage for account status. No two tools overlap in function.
Tool names are lowercase with underscores and mostly follow a verb_noun pattern (list_voices, list_languages, get_usage), but synthesize is a bare verb, which is a minor deviation from the pattern.
With four tools, the server is well-scoped: one core operation, two metadata lookups, and one usage check. This is an ideal size for a TTS-focused server.
The tool surface covers the essential lifecycle for text-to-speech: the synthesis action, necessary resource discovery (voices/languages), and account monitoring. There are no obvious gaps or dead ends.