LinkedIn · talking head video
A LinkedIn video of you, never filmed
To make a LinkedIn video with AI, clone your face from one photo and your voice from a short recording, write about twenty seconds, and render it square or 4:5 so it fills the feed. Feed video usually starts muted, so the first line of the post and the captions you add have to carry the point. On Percify using your own photo and saving your cloned voice cost nothing, and a 20 second clip at 480p is 40 credits on the fast model (2 a second).
- 1:1, 4:5
- the shapes that fill the feed
- 20 s
- a good length for one idea
- 0
- credits to set up your face and voice
- 40
- credits, a 20 second fast clip
Credits cost about 2 cents each depending on the plan, so a minute of talking video at 120 credits works out at roughly $2 to $3.
Why it should be you, not a presenter
LinkedIn is a feed of people, and the audience follows a person. A stock presenter reads as an advert; your own face in your own voice reads as a post. Clone yourself once from one photo and a 15 to 30 second recording, and every clip after that is you. Setup is in clone yourself.
Shape and length for the feed
Square and 4:5 clips take up more of the feed than a wide video, which shrinks to a strip. For length, aim for about twenty seconds. On LinkedIn that is one idea, said well, rather than a talk.
Muted by default: write for the first line
Feed video usually plays silently until someone taps it, so the post text above the video and the captions on it do the work in the first seconds. Put the point in the first line of the post, add captions in your editor before posting, and open the clip on the idea rather than on hello.
Ideas that suit a twenty second clip
Short clips reward one clear thought. A few that work for consultants, teachers, agents and founders:
- One lesson from this week, and what you would do differently.
- A common myth in your field, and the correction.
- A number from your work, and what it means for the viewer.
- An announcement, with one reason it matters.
- The answer to a question clients ask you every month.
Posting it honestly
It is your likeness saying your words, so it is yours to post. Decide once how you disclose AI generated video, and be consistent; LinkedIn audiences notice when authenticity is performed rather than kept. The checklist is in ethical AI video generation.
Post one in four steps
- 1
Clone yourself once
One photo for the face, a 15 to 30 second recording for the voice.
- 2
Write one idea
About twenty seconds, opening on the point.
- 3
Render square or 4:5
At 480p on the fast model, 2 credits a second.
- 4
Add captions and post
Captions in your editor, the point in the first line of the post.
Clips that fit the feed
Real Percify output from one photo each, framed for a feed.
Square format
made for the feed
Alex
talks from one photo
Company update
a leadership message
Frequently asked questions
Can I make a LinkedIn video without filming myself?
Yes. Clone your face from one photo and your voice from a short recording, write the script, and render a talking video of yourself. There is no camera step.
What shape should a LinkedIn video be?
Square or 4:5. Both take up more of the feed than a wide 16:9 video.
How long should a LinkedIn video be?
Short. About twenty seconds for one idea.
Does LinkedIn video play with sound?
Feed video usually starts muted, so add captions in your editor and put the point in the first line of the post.
Should I say the video is AI generated?
Decide your position once and keep it. It is your own likeness, and being open about how it was made protects the trust a personal post depends on.
Keep reading
Post a video this week
One photo, your voice, one idea in twenty seconds.