tidal-mcp

MCP音视频

提供方 dragomirweb

音视频 stdio · 本地进程 OAuth 授权 社区 / 第三方 Claude DesktopClaudeCursorZed
查看接入方式

TIDAL MCP is a Model Context Protocol server that connects AI assistants to the TIDAL music streaming API for search, playlist management, recommendations, and library control.

一句话选型结论

按资料字段整理

tidal-mcp 属于音视频类 MCP 服务器,stdio · 本地进程,OAuth 授权,可直接接入 Claude Desktop、Claude、Cursor 等,有可直接执行的安装命令。接入前先确认权限范围;涉及写入、删除或付款的操作,建议先在测试环境验证。

分类音视频 传输方式stdio · 本地进程 鉴权OAuth 授权 部署方式本地部署
GitHub 星标 2 仓库热度,无公开仓库时显示未标注
工具数 未标注 暴露给模型的工具数量
资源 / 提示词 0 / 0 资源与提示词模板数量
维护方 社区 / 第三方 核验等级 仅资料核对

工具与能力

音视频
数量来自来源目录的标注,具体工具清单以官方文档为准
工具数量未标注暴露给模型可调用的工具个数
资源数量0可读取的资源条目
提示词数量0内置提示词模板数量

能力标签

Local Deployment Music Streaming API Integration Python AI Assistant Playlist Management Music Recommendations

支持客户端

适用场景

Local DeploymentMusic StreamingAPI Integration

安装与配置

stdio · 本地进程 OAuth 授权
客户端在本机拉起子进程,数据不出本机
docker run -i --rm -v /Users/YOUR_USERNAME/.tidal-mcp:/app/session-data -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json dragomirweb/tidal-mcp

这条命令只是入口,具体启动参数请按官方文档补齐,不要直接照搬到别的服务器。

  1. 1
    准备运行环境

    stdio 服务在本地起进程,需要先装好 python,再按安装命令拉取依赖。

  2. 2
    安装服务

    复制上面的安装命令执行;命令里带占位符的位置,换成自己的密钥或路径。

  3. 3
    写入客户端配置

    在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,让新配置生效。

  4. 4
    验证是否可用

    连接成功后随便调一个只读工具,确认能返回结果,再接入正式流程。

典型使用场景

3 类
按分类与标签整理的常见用法,具体以官方文档为准
转写与字幕

把音视频转成文字和字幕。

音视频处理

裁剪、转码、拼接素材。

内容分发

把成品发到对应平台。

选择建议

选之前先看这几条
按条目字段给出的判断,不替你决定用不用

值得选的地方

  • 有现成的安装命令,不用自己找入口,照着跑就能起服务。
  • 已适配 Claude Desktop、Claude、Cursor,换客户端不用重新研究配置。

先想清楚的地方

  • stdio 方式要求本地装好运行时(如 Node.js),换机器要重新装一遍。
  • 使用前要准备好OAuth 授权对应的凭据,并按最小权限配置。
  • 资料没标注工具数量,实际能力要装好后确认。
  • 版本更新与行为细节以官方仓库为准。

更适合

  • 需要在音视频场景里让模型直接调用外部能力的人。
  • 已经在用 Claude Desktop、Claude 的用户,接入路径最短。
  • 想先快速试一下、能接受在本地或自有环境跑服务的人。

不太适合

  • 完全不想在本地装环境的人。
  • 不方便管理密钥或第三方授权的人。
  • 需要厂商级 SLA 和技术支持的生产环境。

同类对比

6 个同类
同分类里挑最接近的几条,字段以各自详情页为准
MCP 服务器工具数传输方式部署方式安装命令最后核验
tidal-mcp当前 未标注 stdio · 本地进程 本地部署 有 2026-06-20
Ableton MCP B7b69a6a 16 stdio · 本地进程 本地部署 有 2026-05-17
Mini Max MCP 8 SSE · 服务端推送 远程部署 有 2026-05-18
LocalAI 未标注 stdio · 本地进程 本地部署 有 2026-04-11
Youtube E0eb3510 未标注 stdio · 本地进程 本地部署 有 2026-05-17
Voicemode MCP 未标注 stdio · 本地进程 本地部署 有 2026-05-17

服务器档案

音视频
字段来自项目仓库与公开索引,中文说明由本站标注
提供方dragomirweb上游标注的维护方 dragomirweb
分类音视频本站按上游分类展示
连接方式stdio · 本地进程客户端在本机拉起子进程,数据不出本机
鉴权方式OAuth 授权远程服务建议用最小权限凭据
维护方社区 / 第三方核验等级 仅资料核对
包名 / 运行时未标注 · python本地安装时用到的包标识与运行时
部署类型本地部署上游给出的部署形态
最近核验2026-06-20上游最后核验时间

详细介绍

15.7 KB
来自项目仓库与上游目录的原文,未做删改

TIDAL MCP

A Model Context Protocol (MCP) server that connects AI assistants to the TIDAL music streaming API. Search, manage playlists, get recommendations, and control your TIDAL library through Claude or Cursor.

How It Works```

Claude Desktop / Cursor
│
│ MCP JSON-RPC over stdio
▼
start_mcp.py ← entry point + stdin proxy for clean shutdown
│
mcp_server/server.py ← FastMCP app, 17 tool definitions
│
│ direct Python function calls
▼
tidal_api/routes/ ← business logic (tracks, playlists, search, auth)
│
tidal_api/browser_session.py ← tidalapi.Session wrapper
│
│ TIDAL REST API
▼
TIDAL


Tool calls use a validated OAuth session from disk and keep it cached in-process after the first successful load. Auth tools invalidate that cache when needed. The MCP layer calls the relevant `tidal_api/routes/` function directly and returns. No internal HTTP server, no child processes, no ports.

## Installation

### Prerequisites

- Python 3.10+ **or** Docker
- uv — only required for local (non-Docker) installation
- A TIDAL subscription

### Session directory

All installation methods store the TIDAL OAuth session in a local directory. Create it once before proceeding:

**macOS / Linux:**```bash
mkdir -p ~/.tidal-mcp

Windows (PowerShell):```powershell
mkdir "$env:APPDATA\tidal-mcp"


This gives you a fixed, predictable location (`~/.tidal-mcp/` or `%APPDATA%\tidal-mcp\`) that the MCP client configs below reference by absolute path.

### Option 1: Docker Hub (easiest)

No cloning or building required — pull the pre-built image directly from Docker Hub. Supports both amd64 and arm64 (Apple Silicon).

1. Pull the image:```bash
   docker pull dragomirweb/tidal-mcp
   ```

2. Authenticate with TIDAL:

   **macOS / Linux:**```bash
   docker run -it --rm \
     -v ~/.tidal-mcp:/app/session-data \
     -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json \
     dragomirweb/tidal-mcp \
     .venv/bin/python auth_cli.py
   ```

   **Windows (PowerShell):**```powershell
   docker run -it --rm `
     -v "$env:APPDATA\tidal-mcp:/app/session-data" `
     -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json `
     dragomirweb/tidal-mcp `
     .venv/bin/python auth_cli.py
   ```

   Open the URL shown in the output, log in to TIDAL, and the session is saved automatically.

3. Configure your MCP client (see MCP Client Configuration) and restart it.

### Option 2: Docker (build from source)

1. Clone the repository:```bash
   git clone https://github.com/dragomirweb/tidal-mcp.git
   cd tidal-mcp
   ```

2. Build the image:```bash
   docker build -t dragomirweb/tidal-mcp .
   ```

3. Authenticate with TIDAL:

   **macOS / Linux:**```bash
   docker run -it --rm \
     -v ~/.tidal-mcp:/app/session-data \
     -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json \
     dragomirweb/tidal-mcp \
     .venv/bin/python auth_cli.py
   ```

   **Windows (PowerShell):**```powershell
   docker run -it --rm `
     -v "$env:APPDATA\tidal-mcp:/app/session-data" `
     -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json `
     dragomirweb/tidal-mcp `
     .venv/bin/python auth_cli.py
   ```

   Open the URL shown in the output, log in to TIDAL, and the session is saved automatically.

4. Configure your MCP client (see MCP Client Configuration) and restart it.

### Option 3: Local Python

1. Clone the repository:```bash
   git clone https://github.com/dragomirweb/tidal-mcp.git
   cd tidal-mcp
   ```

2. Install dependencies:```bash
   uv sync
   ```

3. Configure your MCP client (see below) and restart it. Authentication is handled through the `tidal_login` tool — no separate auth step needed.

## MCP Client Configuration

### Claude Desktop

Edit the config file:
- **macOS**: `~/Library/Application Support/Claude/claude_desktop_config.json`
- **Windows**: `%APPDATA%\Claude\claude_desktop_config.json`

**Docker — macOS / Linux:**```json
{
  "mcpServers": {
    "TIDAL Integration": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/YOUR_USERNAME/.tidal-mcp:/app/session-data",
        "-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
        "dragomirweb/tidal-mcp"
      ]
    }
  }
}

Replace /Users/YOUR_USERNAME/.tidal-mcp with the absolute path to your ~/.tidal-mcp/ directory (run echo ~/.tidal-mcp to get it).

Docker — Windows:```json
{
"mcpServers": {
"TIDAL Integration": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"-v", "C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp:/app/session-data",
"-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
"dragomirweb/tidal-mcp"
]
}
}
}


Replace `C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp` with the output of `echo %APPDATA%\tidal-mcp` in a command prompt.

> Docker is the recommended approach on Windows, especially when using WSL. It avoids path and venv compatibility issues between Windows and WSL Python environments.

**Local — macOS / Linux:**```json
{
  "mcpServers": {
    "TIDAL Integration": {
      "command": "/path/to/tidal-mcp/.venv/bin/python",
      "args": ["/path/to/tidal-mcp/start_mcp.py"]
    }
  }
}

Local — Windows (native Python):```json
{
"mcpServers": {
"TIDAL Integration": {
"command": "C:\path\to\tidal-mcp\.venv\Scripts\python.exe",
"args": ["C:\path\to\tidal-mcp\start_mcp.py"]
}
}
}


> Requires a native Windows Python installation (not WSL). Run `uv sync` from a Windows terminal to create the `.venv\Scripts\` venv.

### Cursor

Add to `~/.cursor/mcp.json`. The configuration is the same as Claude Desktop above — use the Docker or local variant that matches your setup:

**Docker — macOS / Linux:**```json
{
  "mcpServers": {
    "TIDAL Integration": {
      "command": "docker",
      "args": [
        "run", "-i", "--rm",
        "-v", "/Users/YOUR_USERNAME/.tidal-mcp:/app/session-data",
        "-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
        "dragomirweb/tidal-mcp"
      ]
    }
  }
}

Docker — Windows:```json
{
"mcpServers": {
"TIDAL Integration": {
"command": "docker",
"args": [
"run", "-i", "--rm",
"-v", "C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp:/app/session-data",
"-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
"dragomirweb/tidal-mcp"
]
}
}
}


**Local:**```json
{
  "mcpServers": {
    "TIDAL Integration": {
      "command": "/path/to/tidal-mcp/.venv/bin/python",
      "args": ["/path/to/tidal-mcp/start_mcp.py"]
    }
  }
}

OpenCode

Add to opencode.json in your project root, or ~/.config/opencode/opencode.json for global access. See the OpenCode MCP docs for details.

Docker — macOS / Linux:```json
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"tidal": {
"type": "local",
"command": [
"docker", "run", "-i", "--rm",
"-v", "/Users/YOUR_USERNAME/.tidal-mcp:/app/session-data",
"-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
"dragomirweb/tidal-mcp"
]
}
}
}


Replace `/Users/YOUR_USERNAME/.tidal-mcp` with the absolute path to your `~/.tidal-mcp/` directory.

**Docker — Windows:**```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "tidal": {
      "type": "local",
      "command": [
        "docker", "run", "-i", "--rm",
        "-v", "C:\\Users\\YOUR_USERNAME\\AppData\\Roaming\\tidal-mcp:/app/session-data",
        "-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
        "dragomirweb/tidal-mcp"
      ]
    }
  }
}

Replace C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp with the output of echo %APPDATA%\tidal-mcp.

Local — macOS / Linux:```json
{
"$schema": "https://opencode.ai/config.json",
"mcp": {
"tidal": {
"type": "local",
"command": ["/path/to/tidal-mcp/.venv/bin/python", "/path/to/tidal-mcp/start_mcp.py"]
}
}
}


**Local — Windows:**```json
{
  "$schema": "https://opencode.ai/config.json",
  "mcp": {
    "tidal": {
      "type": "local",
      "command": ["C:\\path\\to\\tidal-mcp\\.venv\\Scripts\\python.exe", "C:\\path\\to\\tidal-mcp\\start_mcp.py"]
    }
  }
}

Claude Code

Uses the claude mcp add CLI command. See the Claude Code MCP docs for details.

Docker — macOS / Linux:```bash
claude mcp add --transport stdio --scope user tidal-mcp -- \
docker run -i --rm \
-v /Users/YOUR_USERNAME/.tidal-mcp:/app/session-data \
-e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json \
dragomirweb/tidal-mcp


Replace `/Users/YOUR_USERNAME/.tidal-mcp` with the absolute path to your `~/.tidal-mcp/` directory.

**Docker — Windows:**```bash
claude mcp add --transport stdio --scope user tidal-mcp -- ^
  docker run -i --rm ^
  -v C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp:/app/session-data ^
  -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json ^
  dragomirweb/tidal-mcp

Replace C:\Users\YOUR_USERNAME\AppData\Roaming\tidal-mcp with the output of echo %APPDATA%\tidal-mcp.

Local — macOS / Linux:```bash
claude mcp add --transport stdio --scope user tidal-mcp -- \
/path/to/tidal-mcp/.venv/bin/python /path/to/tidal-mcp/start_mcp.py


**Local — Windows:**```bash
claude mcp add --transport stdio --scope user tidal-mcp -- ^
  C:\path\to\tidal-mcp\.venv\Scripts\python.exe C:\path\to\tidal-mcp\start_mcp.py

The commands above use --scope user to make the server available across all projects. The resulting config is stored in ~/.claude.json:```json
{
"mcpServers": {
"tidal-mcp": {
"type": "stdio",
"command": "docker",
"args": [
"run", "-i", "--rm",
"-v", "/Users/YOUR_USERNAME/.tidal-mcp:/app/session-data",
"-e", "TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json",
"dragomirweb/tidal-mcp"
]
}
}
}


> Use `--scope local` (default) for per-project config, or `--scope project` to create a shared `.mcp.json` in the project root.

## Authentication

The first time you use TIDAL MCP (or after your session expires), ask your AI assistant to log you in:

> _"Log me into TIDAL"_

The assistant calls `tidal_login`, which returns a URL immediately. Open the URL in your browser and approve the TIDAL authorization. The assistant then polls `tidal_check_login` automatically until the login completes.

**Docker users** can also pre-authenticate before using the MCP client (see the auth commands in Installation above). The session is saved to `~/.tidal-mcp/tidal-session-oauth.json` (macOS/Linux) or `%APPDATA%\tidal-mcp\tidal-session-oauth.json` (Windows) and persists across container restarts via the volume mount.

## Available Tools

### Authentication

| Tool | Description |
|---|---|
| `tidal_login` | Starts the OAuth device flow. Returns immediately with a URL to open, or `success` if already authenticated. |
| `tidal_check_login` | Polls whether the browser OAuth has been completed. Call after `tidal_login` returns `pending`. |

### Tracks & Recommendations

| Tool | Description |
|---|---|
| `get_favorite_tracks(limit)` | Retrieve your TIDAL favorited tracks, sorted by date added (newest first). |
| `recommend_tracks(track_ids, filter_criteria, limit_per_track, limit_from_favorite)` | Get personalized recommendations via TIDAL's track-radio feature. Seeds from explicit track IDs or auto-fetched favorites. |

### Playlist Management

| Tool | Description |
|---|---|
| `create_tidal_playlist(title, track_ids, description)` | Create a new playlist and populate it with tracks in one call. |
| `get_user_playlists` | List all your playlists, sorted by last updated. |
| `get_playlist_tracks(playlist_id, limit)` | Fetch all tracks from a playlist (auto-paginated — no size limit). |
| `delete_tidal_playlist(playlist_id)` | Delete a playlist permanently. |
| `add_tracks_to_playlist(playlist_id, track_ids)` | Append tracks to an existing playlist. |
| `remove_tracks_from_playlist(playlist_id, track_ids, indices)` | Remove tracks by TIDAL ID or 0-based position index. |
| `update_playlist_metadata(playlist_id, title, description)` | Rename a playlist or update its description. |
| `reorder_playlist_tracks(playlist_id, from_index, to_index)` | Move a track to a different position (0-based indices). |

### Search & Discovery

| Tool | Description |
|---|---|
| `search_tidal(query, search_type, limit)` | Search across all content types: `all`, `tracks`, `albums`, `artists`, `playlists`. |
| `search_tracks(query, limit)` | Search for tracks. |
| `search_albums(query, limit)` | Search for albums. |
| `search_artists(query, limit)` | Search for artists. |
| `search_playlists(query, limit)` | Search for playlists. |

## Environment Variables

| Variable | Default | Description |
|---|---|---|
| `TIDAL_SESSION_FILE` | `{tempdir}/tidal-session-oauth.json` | Override the path to the OAuth session file. Useful when running multiple instances or on systems where the default temp path differs between processes. |

## Development

### Setup```bash
uv sync              # Install dependencies
uv run pytest tests/ -v   # Run the test suite (213 tests)

Testing

The test suite covers all 17 MCP tools, all 19 route functions, and all utility helpers. Tests use unittest.mock to mock tidalapi and mcp at import time -- no TIDAL credentials or network access needed.```bash
uv run pytest tests/ -v # Full suite
uv run pytest tests/test_routes.py -v # Route function tests only
uv run pytest tests/test_server.py -v # MCP tool tests only
uv run pytest tests/test_routes.py::TestComprehensiveSearchHappyPath -v # Single class


| Test file | Tests | Covers |
|---|---|---|
| `test_routes.py` | 112 | All route functions (auth, tracks, playlists, search) |
| `test_server.py` | 65 | All 17 MCP tools, `_call()`, `_get_session()` |
| `test_utils.py` | 27 | `bound_limit`, `fetch_all_items`, `format_track_data` |
| `test_browser_session.py` | 6 | `_ensure_https`, `BrowserSession.login_oauth_start` |
| `test_mcp_utils.py` | 3 | `SESSION_FILE` env var override logic |

### Docker

After making code changes, rebuild the image:```bash
docker build -t dragomirweb/tidal-mcp .

Other useful commands```bash

python -m compileall mcp_server/ tidal_api/ # Syntax-check all source files
uv run python auth_cli.py # Standalone OAuth CLI
uv run python start_mcp.py # Run the MCP server directly


### Project structure```
mcp_server/
  server.py          # FastMCP app, 17 @mcp.tool() definitions, session management
  utils.py           # SESSION_FILE path constant (env var controlled)

tidal_api/
  browser_session.py # BrowserSession (tidalapi.Session wrapper)
  utils.py           # format_track_data, bound_limit, fetch_all_items
  routes/
    auth.py          # handle_login_start, handle_login_poll, check_auth_status
    tracks.py        # get_user_tracks, get_recommendations, get_batch_track_recommendations
    playlists.py     # CRUD: create, get, delete, add/remove tracks, update, reorder
    search.py        # comprehensive_search + 4 type-specific search functions

tests/               # 213 unit tests (mocked deps, no credentials needed)
  conftest.py        # Centralized sys.modules mocking for tidalapi + mcp

License

MIT License

Acknowledgements

Original Projects:
- yuhuacheng/tidal-mcp — Original TIDAL MCP implementation
- ibeal/tidal-mcp — Community fork with direct-call architecture, search, and Docker support

Libraries & Frameworks:
- Model Context Protocol (MCP)
- TIDAL Python API

常见问题

8 问
答案由本站依据条目字段整理,不引用上游原文
tidal-mcp 怎么安装?

在客户端按下面的安装命令接入:docker run -i --rm -v /Users/YOUR_USERNAME/.tidal-mcp:/app/session-data -e TIDAL_SESSION_FILE=/app/session-data/tidal-session-oauth.json dragomirweb/tidal-mcp。装完在客户端配置里引用同一个包或端点。

tidal-mcp 怎么在 Claude Desktop、Claude、Cursor 里配置?

它的传输方式是stdio · 本地进程,在客户端的 MCP 配置里填好启动命令和参数,保存后完全退出客户端再打开,配置才会生效。

tidal-mcp 收费吗?

本站不代收费用,也没有标注具体价格。是否收费取决于服务本身和背后的接口额度,正式使用前按官方说明确认。

tidal-mcp 需要 API Key 吗?

资料标注需要OAuth 授权认证,使用前准备好对应凭据,并按最小权限配置。

tidal-mcp 支持哪些客户端?

资料标注支持 Claude Desktop、Claude、Cursor、Zed;其他客户端只要能配置 MCP Server,也能用同样的命令或地址接入。

tidal-mcp 能提供哪些能力?

资料显示能力标签包括 Local Deployment、Music Streaming、API Integration、Python、AI Assistant、Playlist Management。具体工具清单以官方文档为准。

tidal-mcp 需要什么运行环境?

运行时为 python;属于本地进程,需要在本机装好运行环境。

tidal-mcp 是本地还是远程?

它是本地进程(stdio),由客户端在本机拉起。传输方式为stdio · 本地进程。

热门专题

按标签挑服务器
全部专题 →

服务器信息以作者仓库与官方文档为准;分类、传输方式与鉴权方式的中文说明为中文整理,最近核验 2026-06-20。