CapCut AI features: captions, video tools & free limits
Quick answer
CapCut's useful AI features include automatic captions, text-to-speech, reframing, background cutout, audio cleanup and video creation. Start with the task you need to finish, then check that feature on your device and plan. Being able to preview an AI effect does not always mean you can export it for free.
This guide separates everyday editing from generating new media, explains the export checks that matter, and gives you a practical workflow. Feature documentation was checked on September 12, 2026. We have not measured a universal number of hours saved.
Which CapCut AI feature should you use?
Your task | Feature to look for | What to review before exporting |
|---|---|---|
Add readable subtitles to speech | Auto Captions / Recognise Subtitles | Names, numbers, punctuation, timing and line breaks |
Add narration to a written script | Text to Speech | Pronunciation, emphasis, pacing and selected voice access |
Turn a landscape edit into a vertical version | Auto Reframe | Whether the subject and important text stay inside the crop |
Isolate a person or subject | Auto Cutout | Hair, hands, transparent objects and moving edges |
Clean up a spoken recording | Noise reduction / voice enhancement | Whether cleanup also removes useful speech detail |
Build a first edit from a script | Script to Video / AI video maker | Scene relevance, factual accuracy and replacement footage needed |
Create a new shot from text or an image | AI video generation | Selected model, duration, credits and export restrictions |
These feature families appear in CapCut's AI editing overview, desktop feature page, and AI tools hub. They are a task map, not a promise that every option has the same name, price or availability in every app version.
Editing assistance and AI generation are different workflows
If you already have a recording, captions, reframing and cleanup work on that source material. Review the changes against the original. For a tutorial or testimonial, retaining the speaker's intended meaning matters more than making every pause disappear.
A script-to-video workflow can assemble scenes, narration, music and captions into an initial edit. A model-based generator creates new visual material from a prompt or image. CapCut markets both kinds of workflows through its AI video generator. Read the selected workflow and model in the interface rather than treating every “AI video” button as an identical tool.
For example, a narrated product explainer may need an assembled sequence of approved screenshots. A fictional establishing shot may need newly generated footage. Using the second workflow for the first task can introduce incorrect product details that take longer to repair.
How to add and edit AI captions
CapCut's caption instructions cover web, desktop and mobile. On web or desktop, open a project and find Captions → Auto Captions; select the source language and generate. On mobile, import the clip, open Captions, then choose Auto Captions. Available advanced options vary by version and region.
Before exporting:
- Listen while reading the transcript. Correct product names, people's names, prices and technical terms.
- Review caption timing at cuts and speaker changes. Remove text that belongs to an audio section you deleted.
- Preview at the intended phone size. Keep captions clear of faces, demonstrations and platform interface overlays.
- Check whether you need subtitles burned into the video or a separate subtitle file. Confirm file-export support in your version before doing a long edit.
Automatic transcription is the starting draft. Accurate captions are the deliverable. If your required language, caption format or export is unavailable, compare subtitle generators.
A practical workflow for a short talking-head video
Example brief: turn one recorded explanation into a 30-second vertical clip with corrected captions and a clear next step. This is a suggested workflow, not a timed benchmark.
- Make the first cut. Keep the promise, demonstration and useful conclusion. Check that removing filler does not change the speaker's meaning.
- Review the sound. Compare cleanup with the original while listening through headphones. Keep the version that is easier to understand.
- Reframe. Preview the whole clip after cropping; subjects can move out of frame halfway through a shot.
- Caption the final wording. Proofread after major edits so the text matches the exported audio.
- Add the missing asset. Use an approved screenshot or product image where accuracy matters. Generate a new shot only when it helps explain the subject.
- Check export early. Confirm the actual resolution, frame rate and premium-content requirements before making multiple variants.
If you need a new visual rather than another editing effect, create it with Magic Hour's AI video generator, or animate a prepared asset with image-to-video, then bring the downloaded clip into your edit. Compare current limits and the preview result before buying credits.
Which CapCut AI features are free?
There is no reliable universal free-feature list for every device, region and account. Check the feature's badge, any credit charge and the export panel in the version you actually use.
CapCut's Pro export explanation says a project may require Pro because it contains premium assets, features or export settings, even when those elements were available in preview. A basic project and a project with a paid template can therefore have different export requirements.
To diagnose a paywall, inspect the items the app identifies, then replace or remove the premium elements if you want a free export. Recheck the complete project afterward. Do not assume changing only the resolution will remove restrictions caused by an asset elsewhere in the timeline.
For paid work, verify the license for music, templates and other assets separately from the subscription price. Being able to export a file does not establish permission for every intended use.
When should you use a different tool?
Keep CapCut when it handles your existing footage and delivery requirements comfortably. Look elsewhere when a specific requirement is missing, rather than replacing your whole workflow because another tool also has AI. For timeline and clip-editing options, see our short-form video editing comparison.
- You mainly need new footage: compare text-to-video tools.
- You need control from a product or character image: compare image-to-video generators.
- You need speech generated with the scene: compare video models and platforms with audio.
- You are editing long interviews: use our podcast editing tools guide to compare the workflow you need.
Cloud shadows move slowly across the scene while foreground plants respond to a light breeze. The camera makes one gentle push forward. Preserve the original composition, subjects, and visual style. One continuous shot.
Animate an image for your next edit
Use an approved product or reference image to create a short clip. Check the selected model and credit estimate, then inspect the result before making more variants.
Try Image to VideoFrequently asked questions
Some workflows produce a first assembled video from a script, and others generate a new shot. You still need to check scene relevance, factual claims, captions and the export. “Automatically created” is not the same as ready to publish.
No. CapCut's documentation describes platform differences, including advanced caption and export options. Check the actual device and account you will use for the project.
It depends on the source footage, task and corrections required. Compare the time to a usable exported result for your own repeatable task, including proofreading and repairs. We do not have evidence for a fixed weekly savings claim.
