CapCut: AI features inside the editor itself
ByteDance's video editor has grown into a toolbox with script-to-video, avatars and voice cloning. For short social formats it is the shortest route from idea to finished clip.
This article is an overview based on vendor information and publicly available reporting, not a scored test of our own. Prices and features change particularly fast in this field; the vendor's own terms always take precedence. The tools we have used ourselves for weeks are scored in the AI video generators & voice AI category.
What CapCut is today
CapCut started as a free video editor for the phone and is today above all one thing: the application in which a very large share of all social media video actually gets cut. Since ByteDance added generative features, the AI sits where the work already happens instead of in a second tool.
That now includes script-to-video, automatic captions, speech output, voice cloning and avatars. The appeal lies not in any single feature being best in class, but in editing, audio, text and effects living on the same timeline.
For an editorial team that means: if you regularly produce short formats, you save the export and import trip between several programs. In practice that trip costs more time than the generation itself.
Pricing and the limits of the free tier
The free tier is unusually generous: editing, captions and part of the AI features work without a subscription. You pay for resolution, watermark removal, team features and the compute-heavy generative steps, which are billed through a credit system.
According to the vendor, entry costs around 10 US dollars a month, with the higher tier just under 20. Generative steps such as avatar video or voice cloning consume additional credits, and that consumption is hard to predict.
So do not budget from the subscription price but from credit consumption per finished clip. How to work that out cleanly is covered in our article on free AI tools.
Where CapCut runs out of road
Generative quality sits below that of specialised models. If you need photorealistic clips with deliberate camera work, Runway or the models in our Kling portrait will take you further.
Speech output is usable but not at the level of ElevenLabs, particularly on German names. It is fine for captions and quick voice tracks, not for an audiobook.
The third point is data: CapCut belongs to ByteDance. If you process material involving personal data, check the processing locations before customer material lands on the timeline.
Who it is worth it for
Anyone who needs many short videos in little time and has no post-production department: marketing teams, freelancers, clubs, schools. The entry barrier is low, the learning curve flat, and the free tier goes surprisingly far.
For advertising with ambition, for formats with recurring characters, or for anything destined for broadcast, CapCut is the wrong base. Then the route runs through specialised models and real editing software.
In short: the shortest route to a finished social clip, with a generous free tier. Generative quality and speech output stay behind specialist tools.
Tools discussed in this article
Each tool has a full review with scores and pricing.
Jean-Marc Mihoc works with AI tools every day and tests them here on real tasks, with accounts he pays for himself, rather than on demos. He writes down what comes out of it, even when that is unspectacular, and says for every tool who should not bother with it.
