CapCut introduces prompt-based video edits, with a ChatGPT plugin to follow

Video creators are being offered more ways to direct AI edits: describe a change, select part of a shot, or refine a cut through conversation. At CapCut World in Los Angeles, the company announced editing updates and a planned ChatGPT plugin, with the integration and a separate automation pilot expected in the coming months.

CapCut’s October 27 announcement covers desktop, web and mobile workflows, alongside creator programs and an AI-content transparency pilot. The company says its platform serves more than 700 million monthly users across more than 170 markets and 30 languages, giving the changes a large potential audience.

A conversation that keeps up with manual edits

The planned ChatGPT plugin would connect idea development with the editing process. CapCut says creators will be able to work through an idea in ChatGPT, use CapCut to assemble a first cut, and then ask for revisions in conversation. Its examples include holding a shot longer, making captions easier to read and keeping music underneath a voiceover.

The proposed workflow also allows hands-on editing between requests. CapCut says a creator could make a manual change and resume the conversation with that change understood. That is a specific promise about continuity: the conversational tool is intended to follow the evolving edit, rather than operate only on the instructions that began it.

Separately, CapCut is developing what it calls a more agentic creation experience—software intended to help carry a creative goal through a first draft and subsequent revisions. The company says that pilot will reach desktop, web and mobile in the coming months, while keeping the creator in control.

Describe the change—or point to it

On desktop, the updated EditPilot takes natural-language instructions for complex or repetitive timeline tasks. CapCut gives the example of changing a particular object’s color while leaving the rest of the shot intact. The announced Generative and Interactive Edit tools offer a different route: select what needs changing directly in the content preview, then refine it there.

The web upgrade combines AI image, video and audio generation with tools for organizing and editing those assets. CapCut says AI assets will remain accessible across devices and projects. Interactive Edit, AI Extend and Intelligent Audio are among the features it is bringing to the browser-based editor.

On mobile, AI Lab provides a conversational space for developing ideas and turning photos and clips into projects such as vlogs or social recaps. CapCut says the resulting stories, assets and creative context can move into its multi-track editor. It is also expanding templates that adapt to how much content a user supplies and their preferred style.

Checking AI origins and sharing creative methods

CapCut is piloting an AI transparency tool that lets people upload a file to check whether it was made with help from AI tools in CapCut, Dreamina, Pippit or Hypic. Its stated scope is those four products, not a universal detector for AI-generated media. The company also identifies AI labels, C2PA Content Credentials and invisible watermarks as measures it uses to help identify AI content.

Two creator initiatives

• Skills Marketplace on the web: Creators can share a visual style, repeatable workflow, technique or character-building approach for others to use, adapt and remix. CapCut describes potential rewards, rather than guaranteed earnings.
• Visual Creator Program: The new program targets graphic design, animation, visual design, AI prompts and Skills. It builds on CapCut’s existing Creator Partner Program, and interested creators can apply online.