Tellers MCP is a hosted Model Context Protocol server for AI video editing. It lets compatible agents create and edit video timelines, inspect the result, open a live preview, and export the finished video.

Tellers is the only MCP able to show a live preview of any video editing timeline as the agent edits it.

Endpoint

https://mcp.tellers.ai/

The server uses Streamable HTTP and OAuth authentication. You do not need to install or run anything locally.

Tellers MCP creating a stock-footage video in an agent conversation with the live video preview open beside it
Create a video and review its live Tellers preview beside the agent conversation.

Set up Tellers MCP in Claude

  1. Open Customize → Connectors in Claude.
  2. Select Add custom connector.
  3. Enter Tellers as the name and https://mcp.tellers.ai/ as the URL.
  4. Select Add, then sign in to Tellers and approve access.
  5. Enable Tellers from the Connectors menu in your conversation.

On Team and Enterprise plans, an organization owner must add the connector before members can connect their accounts. See Claude's remote MCP documentation for plan-specific details.

Set up Tellers MCP in Codex

Run:

codex mcp add tellers --url https://mcp.tellers.ai/
codex mcp login tellers

Complete the OAuth flow in your browser, then run codex mcp list to confirm the connection. Inside Codex, use /mcp to view the available Tellers tools.

Codex shares MCP configuration between its CLI and IDE extension on the same host. See the Codex MCP documentation for configuration options.

Codex using Tellers MCP to add a cinematic title while the updated video preview remains visible
Codex adds an animated title through Tellers MCP and opens the updated preview.

Start editing

Once connected, ask your agent to create a video, edit an existing project, or show you the current timeline preview. For the product overview and workflow examples, visit the Tellers MCP landing page.