AI video assistant: it can edit, and it can create
The TutuCut AI does two things: it edits footage you already have from a single sentence, and it builds an animated video from nothing when you have no footage at all. Characters, scenes and text animation are drawn on the spot by the engine in your browser — no queue, no sign-in, no watermark.
If you have footage, say "cut the first 10 seconds, add a title card, put some light music under it". If you have nothing, say "make a 15-second birthday greeting with a character dancing in a park". It does both.
Creating from scratch needs no queue because it is not running a video generation model. What the AI writes is a timeline — which character appears at which second, doing what, with what text — and the engine in your browser draws it.
Anything you can do by hand, the AI can do too. That is not a slogan: a test called manual-parity checks it item by item, and the build fails if something is missing.
Do you already have footage?
Yes — let the AI edit it
FAQ
Does AI video editing cost anything?
Editing, creating and typed subtitles are free and need no account. Only the server-powered parts require signing in: transcribing what is said in a video, generating photorealistic images or voice-over, and 4K cloud rendering. Those come with a daily free allowance.
How long is the queue?
There is none. Animation built from the built-in characters, scenes and text is drawn immediately — usually within seconds. Only speech recognition takes a dozen seconds or so.
I have no footage at all. Can it really make a video?
Yes. Built in are 12 character actions, 8 scenes, 6 props and 6 templates, plus text animation, 26 filters, 13 transitions and 20 effects. The AI assembles an animated video from these without needing a single file from you.
Do AI-generated videos have a watermark?
No. You get a clean MP4 you can publish directly.
Can I undo what the AI did?
Yes. Each AI action counts as one step, so Ctrl+Z works exactly as it does for your own edits.
Does the AI see my video content?
Not by default. It reads the project structure — which clip sits at which second — not the pixels. Only when you explicitly ask it to "listen to this part" or "look at this frame" is that segment sent for analysis, and it lands in your own Tutu Drive.
Are the AI features the same on mobile?
Yes. On a phone the assistant is a bottom sheet, and it can do everything it does on desktop.
What can the AI definitely not do?
It cannot generate live-action footage of real people, cannot restore damaged video quality, does not do super-resolution, cannot remove an existing watermark, cannot identify people, and cannot handle WMV, FLV or RM files. Without a network the AI is unavailable, though local editing keeps working.