Seedance 2.5 is live Limited-time offer: up to 51% off, from just $0.046/sec
Upgrade Now
ChatCut Docs

What is ChatCut?

ChatCut is an AI-driven video editing agent. Tell it what you want in natural language and it assembles or revises the edit and generates new media — then you make the final precise adjustments yourself on a multi-track timeline.

How do I use ChatCut?

Today, ChatCut runs as the Desktop app, the Agent Plugin, or the web app. Pick a surface to see how it works:

ChatCut Desktop

We recommend the desktop app for most editing work: it reads local media straight from your computer, runs ChatCut or your own local agent (Claude Code / Codex CLI) in the same editor, and renders exports on your machine — with 4K and HDR when your hardware allows. Installation and agent connections are covered in Web, Desktop, and Agent Plugin.

Agent Plugin

The conversation stays in the Claude Code, Codex, or WorkBuddy desktop app. After installing the plugin, the agent opens an editable ChatCut project in its own Browser panel and edits from the conversation — and you can step into the editor to adjust by hand at any time.

Start a new conversation in that desktop app, then paste and send the install prompt:

Claude Code

Command

/goal Read chatcut.io/claude to install and use the ChatCut plugin

Codex — search for chatcut in the official OpenAI plugin store, or install by command:

Command

/goal Read chatcut.io/chatgpt to install the ChatCut plugin and set up a new task for me.

WorkBuddy — search for ChatCut under Experts; do not reuse the prompts above.

These prompts only work inside the desktop apps — full details in Web, Desktop, and Agent Plugin.

Web app

Nothing to install: open it in Chrome and sign in. The web app has the full AI panel, My Assets, Viewer, and Timeline, with exports running through ChatCut’s cloud pipeline.

Open app.chatcut.io

What can I make with ChatCut?

Here are some edits worth trying with ChatCut:

Talking head editing

Turn talking-head, interview, or screen-recording footage into a tighter cut: keep the best takes, remove filler words and awkward pauses, then layer motion graphics, music, and captions on top. See Talking Head Edit.

Prompt

Edit my talking head video: find the best takes, remove filler words, cut awkward pauses and stumbles, and tighten the pacing. Make it sound natural and professional. Add style-consistent motion graphics at key moments to emphasize important points. Add background music.

Motion graphics

Generate animated titles, stat cards, charts, and overlays from a text description or image references, then refine them on the canvas. See Motion Graphics.

Prompt

Create motion graphics: [describe what you want to make]

Seedance 2.5 video generation

Generate AI video clips with Seedance 2.5 — text to video, image to video, and reference-guided shots — or edit and extend an existing clip. See AI Video Generation for every model.

Prompt

Generate a video with Seedance 2.5: [describe the video you want]

Product / app promo

Turn an app, product, or landing page into a polished promo. The agent reads the product and its design style from a URL or from assets you upload, writes the script with you, and produces the video step by step. See AI-Generated Ads for finished examples.

Prompt

Create a high-quality promo video for [your product URL or product name]. Ask me for the target duration, then analyze the product, audience, and design style before crafting the video.

AI short film

Plan and produce a cinematic AI short: the agent builds the story, shot list, and per-shot prompts with you, keeps characters and style consistent across shots, then generates the clips and assembles the edit.

Prompt

Create a cinematic AI short film about [story idea or visual concept]. Build the story, shot list, prompts, continuity, and final edit.

Explainer video

Explain any topic with AI narration, stock footage for real-world visuals, and motion graphics for data, diagrams, and abstract concepts. See AI Voiceover.

Prompt

Create a 1-minute explainer video about [topic to explain]. Generate voiceover narration. Use stock footage for real-world visuals, and motion graphics for data, diagrams, or abstract concepts. Add fitting background music.

Collage B-roll

Turn a short spoken line or an abstract concept into animated paper-collage B-roll — cover for the moments in a cut that have nothing concrete to show.

Prompt

Create animated collage B-roll for: [a short spoken line or abstract concept]

Voice cloning

Clone your voice from a short sample, then turn any script into voiceover in that voice and place it on the timeline. See Voice Library / My Voices / Voice Cloning.

Prompt

Create a voiceover using my custom voice: [what the voiceover should say]

Beyond these starting points, you can edit speech by deleting or rearranging words in the Transcript panel, keep alternate versions on multiple timelines, and export finished video, audio, subtitles, graphics, or XML.

How do I start?

Sign in and ChatCut opens a project for you. Add media, tell the agent what you want, review the result, and export it. Start Your First Edit walks through those five steps with screenshots.

Which product you use — Desktop, the web app, or the Agent Plugin — changes where the conversation happens, not the steps. See Web, Desktop, and Agent Plugin to choose one.

How is the editor organized?

  • AI: describe edits and attach references, Skills, styles, voices, or generation models.
  • My Assets / Library / Transcript: import and organize media, browse built-in assets, or edit speech as text.
  • Viewer: play and inspect the frame at the current playhead.
  • Timeline: arrange clips on video and audio tracks and switch between multiple timelines.
  • Top bar: return to projects, rename or share, undo, manage the workspace and versions, and export.

Continue to Start Your First Edit for the walkthrough, Editor Layout for a button-by-button tour, or AI Panel Overview for the AI workflow.