ElevenLabs MCP Server is an MCP server that integrates ElevenLabs' text-to-speech API to generate audio from text. It supports multiple voices and structured script parts, enabling multi-voice voiceover production workflows within the MCP ecosystem. All generated results and metadata are stored in an SQLite database for persistent history storage and easy retrieval. The project also ships a sample web-based MCP Client built with SvelteKit, demonstrating how to perform simple text-to-speech, manage multi-part scripts with different voices, track voiceover history, and download produced audio files. The server exposes tools to list available voices, generate audio with default or scripted settings, delete jobs, and fetch generated audio files. For installation, you can install via Smithery or run with uvx, and configure credentials and defaults using environment variables such as ELEVENLABS_API_KEY, ELEVENLABS_VOICE_ID, ELEVENLABS_MODEL_ID, ELEVENLABS_STABILITY, ELEVENLABS_SIMILARITY_BOOST, ELEVENLABS_STYLE, and ELEVENLABS_OUTPUT_DIR. The accompanying client UI provides an approachable interface to create and manage jobs, view history, and retrieve audio assets.