Creators report InVideo's MCP lets agents control its video editor

• InVideo’s MCP is live, connecting Claude, ChatGPT and custom agents to its video editor, according to CharaspowerAI.
• Image generation, video generation and editing can run from an agent’s own console, according to MayorKingAI’s post and reply.
• Prompt-driven creation leaves an editable timeline: AllaAisling’s editor demo combines generated elements with manual adjustments.

An agent gets its own named cursor, and project edits can branch for review before reaching the main timeline. The project can also leave InVideo as an editable Premiere Pro, Final Cut Pro or DaVinci Resolve project.

InVideo’s September 30 announcement confirms that its Model Context Protocol server connects external AI agents directly to the browser-based editor. It gives them access to:

• Timeline editing.
• Colour correction.
• Audio mixing.
• Image, video and audio generation.

The company also advertises access to more than 200 image, video and audio models inside the editor.

The clever part is shared control: InVideo says the agent’s changes appear live under its named cursor, and the human can work alongside it or take over at any point.

AllaAisling shared a 29-second demonstration made with InVideo’s new free editor.

InVideo also describes a footage-first route on its editor page:

1. Upload video, images and audio into the Media panel.
2. Drag the selected files onto the multitrack timeline.

That gives the workspace an entry point for existing footage alongside generated material.

InVideo describes project versioning in its agent guide:

• A main timeline and branches.
• Full edit history.
• Changes that can be proposed for review before reaching main.

Tagged versions and returning to an earlier state are also listed in the launch announcement.

InVideo says users can export an editable project for continued work in:

• Adobe Premiere Pro.
• Final Cut Pro.
• DaVinci Resolve.

Finished-video export is also available for work completed inside InVideo.

InVideo’s connection guide describes this setup sequence:

1. The creator opens the Remote Agents page.
2. They choose their AI app and copy its setup prompt.
3. They paste that prompt into the agent’s chat and complete any requested InVideo sign-in.
4. They describe the video or edit. The guide’s example is: “Trim the opening clip and add a title.”
5. The agent provides an Editor link where the creator can review the work.

ChatGPT’s custom connector requires Developer mode. The server uses https://mcp-v2.invideo.io/mcp over Streamable HTTP, with InVideo account sign-in required for tool calls, according to the agent setup guide.

Manual editing is free. InVideo’s announcement says AI generation uses InVideo credits, while the connected assistant’s own plan requirements still apply.