The CapCut app has launched an "AI Assistant" feature. You don't need to remember which menu each function is buried in — just say what you need the way you'd chat with a colleague, and the AI Assistant understands the intent and carries out the operations automatically. From importing footage to exporting the final cut, the whole process runs on natural-language interaction.

Feature Coverage

The AI Assistant covers three major categories:

Basic editing (batch operations)

• Common editing operations such as splitting, transitions, and speed changes
• Batch processing is supported, so there's no need to handle clips one by one by hand

Assets

• BGM addition: just say the song name to set the music, e.g. "add Tanya Chua's 'Red High Heels'"
• Calling up and adding stickers and effects

AI features

• Smart copywriting: writes intro copy and adds the voiceover automatically
• One-click video creation: analyzes your footage, picks the BGM automatically, and generates beat-synced cuts
• Smart enhancement: automatically beautifies the video
• Generate assets from a single sentence

How to Use It

Before You Start

• Update the CapCut app to the latest version
• Have your video footage ready

Step 1: Open the AI Assistant

After importing footage, enter the multi-track editing screen and find the assistant icon at the bottom-left of the track area (it's quite prominent). Tap it, then type text or long-press the voice button and speak to issue a command.

Step 2: Give editing commands in natural language

Enter a prompt like "Help me edit a complete celebrity red carpet vlog", and the AI Assistant will automatically do the following:

1. Analyze the footage content and the request
2. Recommend suitable BGM
3. Intelligently generate beat-synced cuts based on the music's rhythm
4. Select highlight moments from the footage
5. Deliver the first cut

The whole process takes less than a minute.

Step 3: Remove watermarks with AI

If the footage contains a logo or watermark, tell the AI Assistant "remove the watermark that appears in the frame". It will jump to the removal screen, where all you have to do is paint over the area by hand.

Note: watermark removal currently supports only footage under 30 seconds and costs 300 credits per use.

Step 4: Generate copy and voiceover

Have the AI Assistant write a piece of intro copy (you can specify the length, e.g. "within 100 characters"), confirm it, tap "Add text and read aloud", and choose a voice you like from the voiceover library — and just like that, a short video with narration is done.

Tip: Generated subtitles come with punctuation by default, which you may need to remove manually from the copy. If the video feels flat, you can have the AI Assistant add BGM to enrich the mood.

Overall Hands-On Impressions

What it does well:

• Accurate natural-language understanding — no need to describe feature names precisely
• Fast first cuts, with results in under a minute
• AI watermark removal handles larger logos cleanly
• Fast copy generation (a dozen or so seconds), usable with light polishing

Things to watch out for:

• Watermark removal doesn't recognize small logos in the top-right corner of the frame very well
• Subtitles come with punctuation and need manual adjustment
• Currently available only in the app

Who It's For

• Casual users unfamiliar with professional editing software: get basic editing done purely with natural language
• Short-video creators who need to turn footage around fast: a big efficiency boost from raw material to finished video
• Content teams that process footage in bulk: batch operations save enormous amounts of repetitive work

Professional editors won't be replaced, but once AI takes over the basic editing work, they can pour more energy into creative planning and narrative pacing.