GPTVideoEditorGPTVideoEditor
About GPTVideoEditor and Its AI Editing Agent
GPTVideoEditor puts a conversational editing agent alongside a browser timeline. The aim is to make a written request produce an edit you can inspect, change and export, while keeping ordinary video processing on your device.
Edit existing footage with an adjustable timeline
The editor works with videos you import. Supported operations include cuts, clip ordering, timed text, volume and mute changes, playback speed, framing and fades through black. The AI agent issues editing commands, and those commands remain part of the project you can review. It does not generate replacement footage for every request.
Choose OpenAI or Anthropic editing models
Quick and Standard use OpenAI language models. Advanced uses an Anthropic model. Each choice plans changes through the same supported command layer. Visual context is limited to optional sparse frame samples, while pause requests can use local silence ranges. A model’s ability to discuss a video is not proof it has viewed every frame.
An independent service with its own plans
GPTVideoEditor is not an official ChatGPT, OpenAI, Anthropic or CapCut product. Its own Whop subscriptions and video credits pay for the AI agent. Shared credentials with GPTPhotoEditor do not merge the video and photo balances. Manual edits and browser exports do not spend AI credits.
The current browser editor beta
Projects are limited to 15 minutes and imported sources to 1 GB. Browser decoder and encoder support, local memory and available file-writing APIs affect what can be processed. Speed changes affect audio pitch; fades pass through black. Automatic transcription, generated video effects, object removal and cloud rendering are not implemented. Keep source files and review the exported result.
Frequently asked questions
Is GPTVideoEditor part of ChatGPT?
No. It is an independent service using AI models to control a browser video editor.
Does choosing Advanced unlock more video effects?
No. All models use the same supported timeline operations and rendering engine. Model choice changes inference and request cost.