Skip to main content
Glama
ElevenLabs MCP Server

ElevenLabs MCP Server

ElevenLabs MCP 服务器

铁匠徽章

与 ElevenLabs 文本转语音 API 集成的模型上下文协议 (MCP) 服务器,具有服务器组件和用于管理语音生成任务的基于 Web 的示例 MCP 客户端 (SvelteKit)。

特征

  • 使用 ElevenLabs API 从文本生成音频

  • 支持多种声音和脚本部分

  • 用于持久历史记录存储的 SQLite 数据库

  • 示例 SvelteKit MCP 客户端:

    • 简单的文本到语音转换

    • 多部分脚本管理

    • 语音历史追踪与回放

    • 音频文件下载

Related MCP server: ElevenLabs MCP Server

安装

通过 Smithery 安装

要通过Smithery自动为 Claude Desktop 安装 ElevenLabs MCP 服务器:

npx -y @smithery/cli install elevenlabs-mcp-server --client claude

使用 uvx(推荐)

使用uvx时,不需要进行特殊安装。

将以下配置添加到您的 MCP 设置文件(例如,Claude Desktop 的cline_mcp_settings.json ):

{
  "mcpServers": {
    "elevenlabs": {
      "command": "uvx",
      "args": ["elevenlabs-mcp-server"],
      "env": {
        "ELEVENLABS_API_KEY": "your-api-key",
        "ELEVENLABS_VOICE_ID": "your-voice-id",
        "ELEVENLABS_MODEL_ID": "eleven_flash_v2",
        "ELEVENLABS_STABILITY": "0.5",
        "ELEVENLABS_SIMILARITY_BOOST": "0.75",
        "ELEVENLABS_STYLE": "0.1",
        "ELEVENLABS_OUTPUT_DIR": "output"
      }
    }
  }
}

开发安装

  1. 克隆此存储库

  2. 安装依赖项:

    uv venv
  3. 将.env.example复制到.env并填写您的 ElevenLabs 凭证

{
  "mcpServers": {
    "elevenlabs": {
      "command": "uv",
      "args": [
        "--directory",
        "path/to/elevenlabs-mcp-server",
        "run",
        "elevenlabs-mcp-server"
      ],
      "env": {
        "ELEVENLABS_API_KEY": "your-api-key",
        "ELEVENLABS_VOICE_ID": "your-voice-id",
        "ELEVENLABS_MODEL_ID": "eleven_flash_v2",
        "ELEVENLABS_STABILITY": "0.5",
        "ELEVENLABS_SIMILARITY_BOOST": "0.75",
        "ELEVENLABS_STYLE": "0.1",
        "ELEVENLABS_OUTPUT_DIR": "output"
      }
    }
  }
}

使用示例 SvelteKit MCP 客户端

  1. 导航到 Web UI 目录:

    cd clients/web-ui
  2. 安装依赖项:

    pnpm install
  3. 将.env.example复制到.env并根据需要进行配置

  4. 运行 Web UI:

    pnpm dev
  5. 在浏览器中打开http://localhost:5174

可用工具

  • generate_audio_simple :使用默认语音设置从纯文本生成音频

  • generate_audio_script :从具有多个声音和演员的结构化脚本生成音频

  • delete_job :通过 ID 删除作业

  • get_audio_file :通过 ID 获取音频文件

  • list_voices :列出所有可用的声音

  • get_voiceover_history :获取配音任务历史记录。(可选)指定特定任务的任务 ID。

可用资源

  • voiceover://history/{job_id} :通过 ID 获取音频文件

  • voiceover://voices :列出所有可用的声音

执照

该项目根据 MIT 许可证获得许可 - 有关详细信息,请参阅LICENSE文件。

Available Tools

6 tools
delete_jobB

Delete a voiceover job and its associated files

Inputs: job_id (required).

generate_audio_scriptB

Generate audio from a structured script with multiple voices and actors. Accepts either: 1. Plain text string 2. JSON string with format: { "script": [ { "text": "Text to speak", "voice_id": "optional-voice-id", "actor": "optional-actor-name" }, ... ] }

Inputs: script (required).

generate_audio_simpleC

Generate audio from plain text using default voice settings

Inputs: text (required), voice_id.

get_audio_fileC

Get the audio file content for a specific job

Inputs: job_id (required).

get_voiceover_historyB

Get voiceover job history. Optionally specify a job ID for a specific job.

Inputs: job_id.

list_voicesB

Get a list of all available ElevenLabs voices with metadata

Inputs: no named parameters.

Tool Schema Changelog

Recent tool additions, removals, and schema changes observed during successful MCP inspections.

Loading tool schema changelog…

    TDQS

    A3.6/5.0

    Scored across 6 tools

    Disambiguation5/5

    Each tool has a clearly distinct purpose with no ambiguity. delete_job removes jobs, generate_audio_script creates multi-voice audio from structured input, generate_audio_simple creates basic audio, get_audio_file retrieves audio content, get_voiceover_history accesses job history, and list_voices lists available voices. The two generation tools are clearly differentiated by complexity level.

    Naming Consistency5/5

    All tools follow a consistent verb_noun pattern throughout: delete_job, generate_audio_script, generate_audio_simple, get_audio_file, get_voiceover_history, list_voices. The naming convention is perfectly uniform with clear action-object relationships and no mixing of styles.

    Tool Count5/5

    Six tools is an ideal number for this audio generation domain. It provides complete coverage of core workflows (generation, retrieval, management, discovery) without being overwhelming. Each tool earns its place with clear utility in the voiceover job lifecycle.

    Completeness5/5

    The toolset provides complete CRUD/lifecycle coverage for ElevenLabs voiceover operations: create (two generation tools), read (get_audio_file, get_voiceover_history, list_voices), delete (delete_job). There are no obvious gaps - agents can create audio, retrieve results, manage jobs, and discover available voices.

    Maintenance

    ActivityMaintained
    ResponsivenessOutdated data

    Related MCP Connectors

    Related MCP Servers