gandr-mcp (io.github.Gandr-AI/gandr-mcp) — Model Context Protocol Text-to-Speech Server

The gandr-mcp server provides text-to-speech capabilities to MCP-compatible clients. It can synthesize speech from input text and exposes tools to list available voices and supported languages, plus a usage endpoint reporting tokens used by the configured key.

🛠️ Key Features

• synthesize renders text to speech as WAV and base64 output
• Supports 23 languages and six voices
• synthesize includes voice, language, sample rate (8000–48000), pitch, and pacing controls
• Every render is watermarked
• list_voices returns the six available voices
• list_languages returns the 23 supported languages and their codes
• get_usage reports configured-key tokens (one token = one character)

🚀 Use Cases

• Generate spoken audio from text within Claude Desktop, Cursor, and other MCP-compatible agents
• Select language/voice options programmatically before synthesis

⚡ Developer Benefits

• Standard MCP tool interface for synthesize, list_voices, list_languages, and get_usage
• Clear synthesis controls: sample rate, pitch, pacing, voice, and language
• Usage visibility via token counts tied to the configured key

⚠️ Limitations

• Output is watermarked for every render
• Synthesis audio constraints are tied to supported voices/languages exposed by the server