Speak AI

<p align="center">

</p>

<h1 align="center">Speak AI MCP Server & CLI</h1>

<p align="center">
Connect Claude, Cursor, Windsurf, and other AI assistants to your <a href="https://speakai.co">Speak AI</a> workspace.<br/>
83 tools, 5 resources, 3 prompts, 28 CLI commands — transcribe, analyze, search, and manage media at scale.
</p>

<p align="center">
<a href="https://www.npmjs.com/package/@speakai/mcp-server"></a>
<a href="https://modelcontextprotocol.io"></a>
<a href="https://opensource.org/licenses/MIT"></a>
</p>

---

Quick Start

1. Get Your API Key

1. Go to Speak AI API Keys
2. Copy your API Key

2. Choose How to Use It

| Method | Best for |
|---|---|
| CLI | Scripts, pipelines, quick one-off tasks |
| MCP Server | AI assistants (Claude, Cursor, Windsurf, VS Code) |
| Both | Full power — agents orchestrate, CLI automates |

---

Setup

Auto-Setup (Recommended)

Detects installed MCP clients and configures them automatically:

npm install -g @speakai/mcp-server
speakai-mcp init

Manual Setup

<details>
<summary>Claude Desktop</summary>

Add to ~/Library/Application Support/Claude/claude_desktop_config.json (macOS) or %APPDATA%\Claude\claude_desktop_config.json (Windows):

{
  "mcpServers": {
    "speak-ai": {
      "command": "npx",
      "args": ["-y", "@speakai/mcp-server"],
      "env": {
        "SPEAK_API_KEY": "your-api-key"
      }
    }
  }
}

</details>

<details>
<summary>Claude Code</summary>

export SPEAK_API_KEY="your-api-key"
claude mcp add speak-ai -- npx -y @speakai/mcp-server

</details>

<details>
<summary>Cursor</summary>

Add to ~/.cursor/mcp.json:

{
  "mcpServers": {
    "speak-ai": {
      "command": "npx",
      "args": ["-y", "@speakai/mcp-server"],
      "env": {
        "SPEAK_API_KEY": "your-api-key"
      }
    }
  }
}

</details>

<details>
<summary>Windsurf</summary>

Add to ~/.windsurf/mcp.json:

{
  "mcpServers": {
    "speak-ai": {
      "command": "npx",
      "args": ["-y", "@speakai/mcp-server"],
      "env": {
        "SPEAK_API_KEY": "your-api-key"
      }
    }
  }
}

</details>

<details>
<summary>VS Code</summary>

Add to ~/.vscode/mcp.json:

{
  "mcpServers": {
    "speak-ai": {
      "command": "npx",
      "args": ["-y", "@speakai/mcp-server"],
      "env": {
        "SPEAK_API_KEY": "your-api-key"
      }
    }
  }
}

</details>

<details>
<summary>Any MCP Client (STDIO)</summary>

SPEAK_API_KEY=your-key npx @speakai/mcp-server

</details>

Environment Variables

| Variable | Required | Default | Description |
|---|---|---|---|
| SPEAK_API_KEY | Yes | -- | Your Speak AI API key |
| SPEAK_ACCESS_TOKEN | No | Auto-managed | JWT access token (auto-fetched and refreshed) |
| SPEAK_BASE_URL | No | https://api.speakai.co | API base URL |

---

MCP Tools (83)

<details>
<summary>Media (16 tools)</summary>

| Tool | Description |
|---|---|
| get_signed_upload_url | Get a pre-signed S3 URL for direct file upload |
| upload_media | Upload media from a public URL for transcription |
| upload_local_file | Upload a local file directly from disk |
| upload_and_analyze | Upload, wait for processing, return transcript + insights in one call |
| list_media | List and search media files with filters, pagination, and optional inline data (transcripts, speakers, keywords) via include param |
| get_media_insights | Get AI insights — topics, sentiment, summaries, action items |
| get_transcript | Get full transcript with speaker labels and timestamps |
| get_captions | Get subtitle-formatted captions for a media file |
| update_transcript_speakers | Rename speaker labels in a transcript |
| bulk_update_transcript_speakers | Rename speaker labels across multiple media files in one call (max 500) |
| get_media_status | Check processing status (pending -> processed) |
| update_media_metadata | Update name, description, tags, or folder |
| delete_media | Permanently delete a media file |
| toggle_media_favorite | Mark or unmark media as a favorite |
| reanalyze_media | Re-run AI analysis with latest models |
| bulk_move_media | Move multiple media files to a folder in one call |

</details>

<details>
<summary>Magic Prompt / AI Chat (12 tools)</summary>

| Tool | Description |
|---|---|
| ask_magic_prompt | Ask AI questions about media, folders, or your whole workspace |
| retry_magic_prompt | Retry a failed or incomplete AI response |
| get_chat_history | List recent Magic Prompt conversations |
| get_chat_messages | Get full message history for conversations |
| delete_chat_message | Delete a specific chat message |
| list_prompts | List available AI prompt templates |
| get_favorite_prompts | Get all favorited prompts and answers |
| toggle_prompt_favorite | Mark or unmark a chat message as favorite |
| update_chat_title | Rename a chat conversation |
| submit_chat_feedback | Rate a chat response (thumbs up/down) |
| get_chat_statistics | Get Magic Prompt usage statistics |
| export_chat_answer | Export a conversation or answer |

</details>

<details>
<summary>Folders & Views (11 tools)</summary>

| Tool | Description |
|---|---|
| list_folders | List all folders with pagination and sorting |
| get_folder_info | Get folder details and contents |
| create_folder | Create a new folder |
| clone_folder | Duplicate a folder and its contents |
| update_folder | Rename or update a folder |
| delete_folder | Delete a folder (media is preserved) |
| get_all_folder_views | List all saved views across folders |
| get_folder_views | List views for a specific folder |
| create_folder_view | Create a saved view with custom filters |
| update_folder_view | Update a saved view |
| clone_folder_view | Duplicate a view |

</details>

<details>
<summary>Recorder / Survey (10 tools)</summary>

| Tool | Description |
|---|---|
| create_recorder | Create a new recorder or survey |
| list_recorders | List all recorders |
| get_recorder_info | Get recorder details and questions |
| clone_recorder | Duplicate a recorder |
| get_recorder_recordings | List all submissions |
| generate_recorder_url | Get a shareable public URL |
| update_recorder_settings | Update branding and permissions |
| update_recorder_questions | Update survey questions |
| check_recorder_status | Check if recorder is active |
| delete_recorder | Delete a recorder |

</details>

<details>
<summary>Automations (5 tools)</summary>

| Tool | Description |
|---|---|
| list_automations | List automation rules |
| get_automation | Get automation details |
| create_automation | Create an automation rule |
| update_automation | Update an automation |
| toggle_automation_status | Enable or disable an automation |

</details>

<details>
<summary>Clips (4 tools)</summary>

| Tool | Description |
|---|---|
| create_clip | Create a highlight clip from time ranges across media files |
| get_clips | List clips or get a specific clip with download URL |
| update_clip | Update clip title, description, or tags |
| delete_clip | Permanently delete a clip |

</details>

<details>
<summary>Custom Fields (4 tools)</summary>

| Tool | Description |
|---|---|
| list_fields | List all custom fields |
| create_field | Create a custom field |
| update_field | Update a custom field |
| update_multiple_fields | Batch update multiple fields |

</details>

<details>
<summary>Webhooks (4 tools)</summary>

…