Transcribe Audio and Video for LLM Interaction
MCP server that connects AI assistants like Claude to Transcribe.com for automated audio-to-text transcription with speaker separation and 100+ language
Why it matters
Enable LLMs to understand and interact with the content of audio and video files by providing fast, high-quality transcriptions. This asset bridges the gap between spoken word and text-based AI.
Outcomes
What it gets done
Transcribe audio/video files and voice notes quickly.
Extract text content with word-level timestamps and speaker separation.
Store and manage transcriptions via Transcribe.com.
Facilitate LLM interaction with transcribed content.
Install
Add it to your toolbox
Run in your project directory:
curl -fsSL https://spark.entire.vc/get/vb-transcribe | bash Capabilities
Tools your agent gets
Converts audio to text and returns the text immediately using temporary credits.
Returns your account balance.
Returns the content of completed transcriptions with optional filtering and search.
Renames or deletes transcriptions in your Transcribe.com account.
Overview
Transcribe MCP Server
What it does
What it does
Transcribe MCP Server connects AI assistants (Claude, Windsurf, Cursor) to your Transcribe.com account, enabling them to automate audio transcription tasks. The server provides tools for converting audio to text, managing transcriptions, and checking account balance - all through a lightweight integration that returns results in seconds without requiring local ASR models or complex Python dependencies.
When to use - and when NOT to
Use this connector when you need AI assistants to transcribe audio files automatically, whether from local storage or public URLs. It's ideal for processing noisy audio, multi-language content (100+ languages supported), or recordings requiring speaker separation and word-level timestamps. The Local Server variant handles local file paths directly, while the remote server works with public-accessible URLs.
Do NOT use this if you need offline transcription without internet connectivity, as the service requires connection to Transcribe.com. Also avoid this if you're working with audio that cannot be uploaded to a cloud service due to strict data residency requirements.
Inputs and outputs
You provide a private MCP integration URL (obtained from the Transcribe.com online editor Automation popup) and audio files via either local file paths or public URLs. The server accepts these through four tools: convert-to-text for immediate transcription, get-balance for account credits, read-transcriptions for retrieving completed transcriptions with optional filtering, and update-transcription for renaming or deleting transcriptions.
You receive plain text transcriptions with word-level timestamps and speaker separation, account balance information, and transcription management capabilities - all returned directly to your AI assistant.
Integrations
For Claude Desktop, download the pre-built MCP Bundle (MCPB) with one-click installation, automatic dependency management, and built-in error handling. For other assistants (Windsurf, Cursor), configure manually using Node.js and NPX:
{
"key": "transcribe-local",
"command": "npx",
"args": [
"args": ["-y", "github:transcribe-app/mcp-transcribe"]
],
"env": {
"MCP_INTEGRATION_URL": "<your-MCP-integration-URL>"
}
}
The connector supports cloud storage for audio notes and records, and Transcribe.com offers a teams feature for collaboration.
Who it's for
This server is built for users who want AI assistants to handle transcription workflows automatically - researchers processing interview recordings, content creators managing podcast transcripts, or teams collaborating on audio documentation. It suits users who prefer fast, cloud-based transcription over managing local ASR models and Python environments.
How it connects
2024-05-15T00:00:00.000Z
Source README
Transcribe MCP
Automate your transcriptions with AI.
Website
About
Transcribe MCP instantly connects your account to assistants like Claude, Windsurf, Cursor, and more so they can automate tasks on your behalf. The Local Server can add local files for transcription and return result to your Assistant in seconds.
Features
- ⚡ Fast, lightweight and LLM-friendly. No special ASR models needed, no setup and fighting python packages, results in seconds.
- 🏆 High-quality transcriptions. Works with noisy audio, over 100 languages supported, featuring word-level timestamps and speaker separation.
- 🔋 Wide variety of formats out-of-the-box and Cloud storage for your audio notes and records.
- 👥 Collaboration support via Transcribe.com teams feature
🚀 Local installation: Claude Desktop
🔹 Get your private MCP integration URL
- Sign in to the Transcribe online editor
- Copy your private URL from Automation popup
🔹 Download pre-built MCP Bundle (MCPB): Click Here
📦 The MCPB includes:
- ✅ One-click installation in Claude Desktop
- ✅ Secure environment variable configuration
- ✅ Automatic dependency management
- ✅ Built-in error handling and debugging
- ✅ The MCPB file is automatically updated for the latest features and fixes
Option 1: Double-Click Installation (Recommended)
- Double-click the .mcpb file
- Claude Desktop will automatically install the extension
- Follow the configuration prompts
Option 2: Manual Installation
- Open Claude Desktop
- Go to Settings → Extensions
- Click "Install Extension" and select the .mcpb file
Configuration
During installation, you'll be prompted to configure MCP integration URL. Paste your private URL.
🚀 Local installation: Other assistants
🔹 Get your private MCP integration URL
- Sign in to the Transcribe online editor
- Copy your private URL from Automation popup
🔹 Before installing the server, ensure you have Node.js and Git
- Download from: https://nodejs.org/
- Verify Node with: node --version
- Verify NPX with: npx --version
🔹 Add Local Server via your assistant settings and use this snippet for server setup:
{
"key": "transcribe-local",
"command": "npx",
"args": [
"args": ["-y", "github:transcribe-app/mcp-transcribe"],
],
"env": {
"MCP_INTEGRATION_URL": "<your-MCP-integration-URL>"
}
}
Configuration
Replace <your-MCP-integration-URL> with your private URL.
⚡ Available Tools
List of tools is expanded with each new version.
1. convert-to-text: converts audio to text and returns the text immediately.
- Note: This tool use your time credits.
- Note: Remote server expect public-accessible URL, Local Server can use both URL and a path to local file in your file system.
2. get-balance: returns balance of your account.
3. read-transcriptions: returns content of ready transcriptions with optional filtering/search.
4. update-transcription: renames or deletes transcriptions in your account at Transcribe.com
💡 Security Notes
- Private URL: Store your private URL securely, do not share it with others
⚠️ Troubleshooting
"MCP_INTEGRATION_URL environment variable is required"
- Sign in to the Transcribe online editor
- Copy your private URL from Automation popup
- Add it as server's "MCP_INTEGRATION_URL" environment variable in your AI assistant settings
Extension Won't Install
- Ensure you have the latest Claude Desktop version
- Check that
nodeis installed and in your PATH
Discussion
Questions & comments · 0
Sign In Sign in to leave a comment.