Click on "Install Server".
Wait a few minutes for the server to deploy. Once ready, it will show a "Started" state.
In the chat, type @ followed by the MCP server name and your instructions, e.g., "@Descript Complete transcribe this audio and remove filler words"
That's it! The server will respond to your query, and you can continue using it as needed.
Here is a step-by-step guide with screenshots.
Stop manually transcribing audio. Upload recordings, get transcripts, remove filler words, and export edited content — all from your AI assistant.
This MCP server gives your AI assistant full control over Descript: transcription, editing, AI voice (Overdub), collaboration, and export — all via natural language.
Related MCP server: AudacityMCP
Tool | Category | Description |
| Transcription | Submit an audio URL for transcription |
| Transcription | Submit a video URL for transcription |
| Transcription | Fetch transcript by share slug (text, JSON, or VTT) |
| Transcription | List recent transcription jobs |
| Projects | Create a new Descript project |
| Projects | List all projects in your account |
| Projects | Get details for a specific project |
| Projects | Delete a project permanently |
| Media | Upload media to a project via URL |
| Media | Check upload/processing job status |
| Editing | Remove um, uh, like, etc. with AI |
| Editing | Remove silent gaps automatically |
| Editing | Edit content via natural language instruction |
| Export | Export transcript as plain text |
| Export | Export transcript as SRT subtitles |
| Export | Render/export the video |
| Overdub | Generate speech with AI voice |
| Overdub | List available AI voices |
| Collaboration | Get project collaborators |
| Collaboration | Invite a collaborator by email |
Get your Descript API key from your Descript account settings
Set DESCRIPT_API_KEY=your_key_here in your environment
Start the server: npm start
Connect your MCP client to http://localhost:8080/mcp
Transcribe a podcast episode:
transcribe_audio({ url: "https://example.com/episode-42.mp3", project_name: "Episode 42" })
// Returns job_id — poll with get_upload_status until state = "stopped"
// Then get_transcript({ slug: "your-share-slug", format: "text" })Clean up an interview recording:
remove_filler_words({ project_id: "abc123" })
remove_silence({ project_id: "abc123", threshold_seconds: 0.8 })
export_transcript_txt({ slug: "your-share-slug", include_speakers: true })Share for review:
share_project({ project_id: "abc123", email: "editor@example.com", role: "editor" })Variable | Required | Description |
| Yes | Your Descript API key |
| No | Server port (default: 8080) |
GET http://localhost:8080/health
# {"status":"ok","server":"descript-complete","version":"1.0.0"}Built with the MCPize framework. Powered by mastermindshq.business.
This server cannot be installed
Unclaimed servers have limited discoverability.
If you are the server author, to access and configure the admin panel.
Search recordings, summarize meetings, create clips, and automate workflows from your AI assistant.
Clip videos into captioned shorts, add captions & schedule posts — from Claude, ChatGPT, Cursor.
Manage projects, tasks, time tracking, and team collaboration through natural language.
We provide all the information about MCP servers via our MCP API.
curl -X GET 'https://glama.ai/api/mcp/v1/servers/josephtandle/descript-complete'
If you have feedback or need assistance with the MCP directory API, please join our Discord server