Connect DistilBook to Claude

DistilBook speaks MCP (Model Context Protocol). Connect it to Claude, Cursor, or any MCP client and turn documents into narrated explainer videos, with your AI reviewing every script before it becomes audio.

1. Connect: one URL, nothing to install

DistilBook is a hosted MCP server. Add it by URL and sign in with your DistilBook account when prompted:

https://mcp.distilbook.com/mcp

claude.ai / Claude Desktop: Settings → Connectors → Add custom connector → paste the URL. You'll be redirected here to log in and approve access.

Claude Code (terminal):

claude mcp add --transport http distilbook https://mcp.distilbook.com/mcp

Cursor / other MCP clients: add a remote server with that URL; OAuth sign-in is handled automatically.

Alternative: API keys (for scripts & CI)

Prefer a long-lived credential over the login flow? Create a key on the API Keys page and send it as a header instead:

claude mcp add --transport http distilbook https://mcp.distilbook.com/mcp \
  --header "Authorization: Bearer dbk_your_key_here"

Keys start with dbk_, are shown once, and grant full access to your videos and credits, so treat them like passwords.

2. Try it

Some things to say once connected:

  • “Turn this article into a 3-slide video for beginners: <url>, and tighten the narration before approving.”
  • “Upload ~/reports/q3.pdf to DistilBook and make a 2-minute video for executives.”
  • “Is my video done? When it is, render it in 1080p and give me the download link.”
  • “List my videos and publish the latest one.”

Available tools

create_videoStart a video from an article URL or an uploaded document
request_uploadGet an upload slot for a PDF, DOCX, Markdown, or TXT file
get_video_statusLive progress for a job, the polling tool
get_narration / set_narrationReview and edit every slide's script before generation
regenerate_narrationAI-rewrite one slide or all of them
approve_narrationApprove scripts and generate audio + visuals
edit_slide_image / proceed_slideAI region edits on slide images (edit_images jobs)
render_video / get_render_statusExport the final MP4
get_download_linkPresigned MP4 download URL
publish_videoPublic share link
list_videos / get_accountYour library, credits, and plan limits
cancel_video / delete_videoStop or remove a job

Good to know

  • Credits are consumed per slide, exactly as in the web app, and your AI can check get_account before starting.
  • Every video pauses for narration review before audio is generated; nothing is spent on voicing a script you haven't approved.
  • Jobs keep running server-side if your AI disconnects, so progress is never lost.
  • Deleting a video always requires an explicit confirmation.
  • Revoke a key anytime from the API Keys page.