Canviz

AI Producer

One switch turns raw talking-head footage into a produced video

For anyone who records themselves talking to a camera, on a Pro plan. Upload one take of up to 6 minutes and get it back tightened, captioned, titled and cut with stock B-roll picked for what you say.

Free on your first video. No credit card required. See it in the demo.

How it works

AI Producer, step by step

  1. 1

    Upload one take

    Paste a YouTube link or upload an MP4, MOV, WebM or MKV, up to 6 minutes. One person talking to the camera, in one take.

  2. 2

    Switch on Producer

    One switch turns on Tighten, B-roll, animated captions, a hook title and leveled audio. Producer is on Pro plans.

  3. 3

    Canviz cuts and adds B-roll

    Pauses and filler words come out. A short cutaway of what is being said drops in on those words, at least 12 seconds apart.

  4. 4

    Get the video

    The summary says what was done, for example “Tightened −34 s · 3 B-roll cutaways”. Download the MP4 or schedule it to TikTok, YouTube or Instagram.

What you get

What comes out, and what you control

  • Tighten

    Pauses and filler words are cut from the speech track on word boundaries, so the video keeps moving. The summary shows how much came out.

  • B-roll for what you say

    Stock footage cutaways, picked for what is being said and checked by AI, drop in on those words at least 12 seconds apart. A moment with nothing that clearly fits gets no cutaway.

  • Animated captions

    Word-timed captions in 23 styles, placed so they never cover a face. Edit any word, change the style or turn them off.

  • Hook title

    A title burned in for the first 3 seconds, up to 12 words, editable.

  • Leveled audio

    The audio is leveled in the same pass, with nothing for you to set.

  • A summary of what changed

    The finished video says what was done, for example “Tightened −34 s · 3 B-roll cutaways”. Tell Canviz in chat if you want something different.

Details

The facts

Input
A YouTube link or an MP4, MOV, WebM or MKV file up to 1 GB; one take of talking-head footage
Length
Up to 6 minutes; a longer video is trimmed first or switched to Clips
B-roll
Stock footage cutaways of 1.5–3 seconds, picked by AI for the words they show, at least 12 seconds apart
Plans
Pro only; Pro includes up to 30 B-roll cutaways a month
Captions
23 word-timed styles, placed so they never cover a face; editable, on or off
Hook title
Burned in for the first 3 seconds, up to 12 words, editable
Output
MP4, 9:16 by default, 1:1 and 16:9 available; download or schedule to TikTok, YouTube and Instagram
Editing
Every project is a chat; tell Canviz what to change and it revises

Got questions?

Producer is on Pro plans only. Pro includes up to 30 B-roll cutaways a month.

From a stock footage library. Canviz finds footage for what is being said at that point, checks that it clearly shows it, and crops it to your format. If nothing fits, that moment gets no cutaway.

Pauses and filler words in the speech track. Cuts fall on word boundaries, using the transcript's word timings in the spoken language, and the summary shows how much was cut.

One take of up to 6 minutes. If the video is longer, Canviz trims it first or switches to Clips, which cuts a long video into several short ones.

Yes. Every project is a chat, so tell Canviz what to change and it revises. Captions can be edited word by word, the hook title is editable and Tighten can be turned off.

On the words they show, at least 12 seconds apart, 1.5 to 3 seconds each. The summary counts them.

9:16 by default, with 1:1 and 16:9 available. The finished MP4 can be downloaded or scheduled to TikTok, YouTube and Instagram.

Still have questions?

Email us and a person will answer, usually the same day.

Get in touch
More tools

Works on the same clips

Whatever Canviz cuts is what it captions, reframes, voices and schedules.

All tools

Try Canviz free on your first video

No credit card required. See pricing for the paid plans.