Twitter Spaces MCP Server

An MCP (Model Context Protocol) server for downloading and transcribing Twitter Spaces using AI.

Features

• ✅ Check Twitter Space availability

• 📥 Download Twitter Spaces

• 🎤 AI-powered transcription using Deepgram

• 📝 Multiple transcript formats (JSON, TXT, Paragraphs, Time-coded, Summary)

• 📋 List all downloaded spaces

• 🔄 Complete download + transcription workflow

Related MCP server: Video Transcriber MCP Server

Tools

check_space_availability

Check if a Twitter Space is available for download.

Parameters:

• space_url (string): Full Twitter Space URL

download_twitter_space

Download a Twitter Space and optionally wait for completion.

Parameters:

• space_url (string): Full Twitter Space URL

• wait_for_completion (boolean): Whether to wait for completion

transcribe_space

Transcribe a downloaded Twitter Space using AI.

Parameters:

• space_id (string): Space ID (extracted from URL)

• wait_for_completion (boolean): Whether to wait for completion

get_transcript

Download transcript in various formats.

Parameters:

• space_id (string): Space ID

• format (string): Format (json, txt, paragraphs, timecoded, summary)

list_spaces

List all downloaded Twitter Spaces.

download_and_transcribe_space

Complete workflow: download and transcribe a space.

Parameters:

• space_url (string): Full Twitter Space URL

Configuration

• apiUrl (required): Base URL of the Twitter Spaces API backend

• timeout (optional): Request timeout in seconds (default: 30)

Usage Examples

1. Check if a space is available:

Use check_space_availability with space_url="https://x.com/i/spaces/1ZkKzYLnWOLxv"

2. Download and transcribe in one step:

Use download_and_transcribe_space with space_url="https://x.com/i/spaces/1ZkKzYLnWOLxv"

3. Get transcript with speaker diarization:

Use get_transcript with space_id="1ZkKzYLnWOLxv" and format="paragraphs"

Code

Backend Requirements

This MCP server requires a compatible Twitter Spaces API backend running with the following endpoints:

• POST /api/download

• GET /api/status/{download_id}

• POST /api/transcribe

• GET /api/transcription/status/{transcription_id}

• GET /api/transcript/{space_id}/download/{format}

• GET /api/spaces

• GET /api/check-space/{space_id}