Eight people, one script, and no camera
One script, eight presenters, no camera. Watch it with the sound on, then read how every shot in it was made and who is really in them. The answer to that last one is nobody.
One script, eight presenters, no camera. Watch it, then read how it was made.
This is one video we made for our own homepage. Nobody in it is a customer and nobody in it is real.
Part of Showcase.
What does an AI generated video actually look like?
You are looking at one. Eight presenters, one script between them, about twenty five seconds. Nobody was filmed. Each shot came from a single still photo and one line of script, rendered on its own, then the eight were cut together in an edit. Every person in it is a synthetic presenter we made.
Who it is for
Anyone deciding whether this is good enough to put their own name next to.
What you need
Nothing to watch it. To make one with your own face: a clear photo of you, about ten seconds of you speaking, and a script.
What it costs
Watching costs nothing. A talking-head video of you is 152 credits a second on the Plus plan and above, with a floor of 120 credits a render. The Free plan starts with 8,400 credits, once.
What it does not do
It does not show you a customer, because nobody in it is one. And the Studio will not put eight people in a row for you in one press. It makes one presenter at a time.
The tool that does the talking part is talking-head video, and the difference between the characters in this film and a real person is set out on what an AI Digital Twin is.
What you are watching
Eight shots, eight people, one sentence each. Two of them share a sentence.
| At | Who | The line |
|---|---|---|
| 0:00 | WhoAn orthodontist, consulting room | The lineYou know you need to keep creating content. |
| 0:02 | WhoA lawyer, office with bookshelves | The lineBut finding time to film is another story. |
| 0:05 | WhoAn HR director, desk by a whiteboard | The lineThere is always something more important that needs your attention. |
| 0:09 | WhoA creator, at home, holding a product | The lineAnd content keeps getting pushed to later. |
| 0:12 | WhoA man on a couch, lamp behind him | The lineSo instead of making more time for filming, |
| 0:17 | WhoAn executive, corner office | The lineyou can make more from the time you already have. |
| 0:19 | WhoA doctor, exam room | The lineStay consistent without adding more to your schedule. |
| 0:21 | WhoA senior executive, office above a city | The lineSo content gets done without taking over your day. |
Two of them split one sentence between them, at 0:12 and 0:17. That was a choice, and it is the only place in the film where a thought carries across a cut.
How it was made
Eight separate renders and an afternoon of picking takes. The rules below are the ones that cost us the most to learn.
The script came first, and it stayed put
Eight lines, written to be said out loud, one line per person. Nothing was improvised and nothing was rewritten to fit a picture. When a line did not survive, we changed the line.
Every face was approved as a still before anything moved
One photograph per person, looked at, argued about, signed off. A face you are unsure about in a still is a face you will hate in motion. Three of the eight were reworked at the still stage and never reached a render in their first version.
One line, one shot, about four seconds
Give a single line more room than it needs and the read fills the space: it repeats the phrase, or it drops in a word that was never in the script. Both happened here. Four seconds is enough for a sentence and not enough to wander.
We read the words back before we trusted a take
Every shot was written out from its own audio and compared to the script, word by word. It is dull and it catches what your ear will not. One line quietly lost its first half because the take opened on a second of silence and ran out of room at the end.
Pauses are the hardest part, and you cannot direct them
The opening line took nine attempts, because of one small word. “You know” reads as an aside or as a plain statement of fact, and every note we wrote made it worse. What worked was rewriting the line. If a sentence has two readings, change the sentence.
Nobody says our name
Read the transcript. It is not in there once. A spoken brand name comes back wrong often enough that we put ours on the screen instead.
Then it was cut
Eight separate files, in script order, hard cuts, no dissolves, trimmed at the head and the tail until the sentences ran into each other properly. That last part is an editing app and a pair of ears, the same as it has always been.
Why it is eight people and not one
Because each shot was made on its own, and a separate render means a separate voice. There is no setting that carries one voice across eight separate pieces, so eight people is not a flourish here. It is the honest shape of what was made.
We leaned into it: eight professions, eight rooms, one sentence each, so the cut reads as a relay rather than as a continuity error.
It also tells you something useful about the tool. Write one script and render it and you get one person saying all of it, in one voice, start to finish. That is the normal case, and it is the thing most people want. This film is the unusual case, assembled by us on purpose.
What is not real here
Nobody in this video exists. All eight are Avatars, which in our vocabulary means a character with no real person behind them: no consent record, because there is nobody to consent, and no likeness to protect. They are not customers, they are not staff, and nothing they say is a testimonial. Two are dressed as clinicians and one as a lawyer, and none of them is any of those things.
The other route is the opposite of this one, and it is the one most people here are actually buying. A real person's photo, their own voice and a consent recording that has to be on file before the first render will run. That gate is not a setting, it is the door.
We also call this a film, which it is not. Nothing was filmed. We kept the word because it is what you call a finished short piece, and pretending otherwise would be silly on a page about how it was put together.
- Digital Twin
- A real person: their photo, their voice and a recorded consent clip, checked against the photo before anything renders. Every face in the film above is the other thing.
What you would get instead
One presenter, saying your whole script, in one voice, start to finish.
You write the words, pick who says them and press render. Free renders run to one minute and at 720p; 1080p is on the Plus plan and above. If you want the finished video to speak to somebody who does not read English, it can be re-voiced afterwards in ten languages with the same voice carrying on.
The product page for the door this film went through, eight times.
Before you publish
Checked September 2026. This is a summary, not legal advice.
The transcript
The same words the caption file carries, in the order you hear them.
Read the transcript
0:00You know you need to keep creating content.
0:02But finding time to film is another story.
0:05There's always something more important that needs your attention.
0:09And content keeps getting pushed to later.
0:12So instead of making more time for filming, you can make more from the time you already have.
0:19Stay consistent without adding more to your schedule.
0:21So content gets done without taking over your day.
Questions about this video
- Is anyone in this video a real person?
- No. All eight are Avatars we made for this film. None of them is a customer, none is an employee, and nothing they say is a testimonial. A real person's video is a different thing: a Digital Twin needs that person's photo, their voice and a consent recording on file before it will render anything at all.
- Why does each person only say one line?
- Two reasons, both practical. A single line given too much room starts to pad itself, so each shot was kept to about four seconds. And because each shot was rendered on its own, each one came back with its own voice, which makes a relay of eight people the honest way to cut them together rather than a stylistic choice.
- Could I make this exact video in the Studio?
- Not in one press. This was assembled by us, shot by shot, and cut together in an editing app afterwards. What one press gives you is one presenter saying your whole script, in one voice, which is what most people want anyway.
- Is it available in other languages?
- This film is English only. Once a video is finished here it can be re-voiced in another language with the same voice carrying on, across ten languages, on the Plus plan and above. Dubbing is the page for that, and it has its own short demo.
- What was the hardest part?
- One word. The opening line starts with “you know”, which reads as a throwaway aside or as a plain statement of fact, and it kept coming back as the aside. Every note we wrote to fix the delivery made it worse or lost the words. The fix was to stop directing the read and rewrite the line.
The tools that do this
The doors this film went through, and the pages that explain them.
Now make one with your own face
You will not get eight people. You will get you, saying whatever you write, as many times as you change your mind about what it should say. The Free plan starts with 8,400 credits, once, and no card to hand over first.