Vizard Agent

How to make your own version of a tutorial reel you liked

Last updated 2026-08-18 · 6 min read

Send Vizard Agent a link to a tutorial reel you liked and say what yours should be about. Vizard Agent downloads the reference, transcribes it, maps every shot in it, analyses the overlays, the pacing and the audio, and then builds your own version to that structure with its own generated character and footage.

What is the short version?

"Can you make something like this" is the most common brief there is and the hardest to answer well, because what you actually admired is almost never the content. It is the structure: how fast it cuts, when the text appears, where the tone changes.

  1. Go to Vizard Agent and paste the link to the reel.
  2. Say what your version should be about.
  3. Say the feed, the length and whether it needs a narrator.

What do you need before you start?

The link and your own subject. Vizard Agent studies the reference on its own, so what it needs from you is the thing the reference cannot supply, which is what your video is actually teaching and who it is for.

What do you type into Vizard Agent?

Paste the link and name your subject. Vizard Agent works out the structure by watching the reference itself, so a long description of what you liked about it mostly repeats what it is about to go and measure, and the sentence that changes the result is the one naming what your version teaches.

Prompt

[link] Can you make something like this, but about [your subject]?

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real reference-led tutorial. The first half is a study of the reel you sent, and the interesting part is what it measures: not just the shots, but the overlays and where the energy changes.

  1. Loads the tutorial playbook and reads the download tool's options.
  2. Downloads the reference reel and probes it.
  3. Extracts survey frames and reviews the contact sheet.
  4. Opens individual frames across the reel at the start, the quarter, the middle and the end.
  5. Transcribes the spoken audio of the reference.
  6. Maps every shot in it, so the structure exists as a list rather than an impression.
  7. Analyses the overlay treatment, the pacing and the audio as separate properties.
  8. Looks at the overlay detail in the calm section and again in the chaotic one.
  9. Generates a character portrait and six candidate music tracks, then builds a matching full-body reference so the presenter stays the same person throughout.

Steps 7 and 8 are the ones worth stealing. Vizard Agent compared how the overlays behave in the quiet part against the busy part, which is exactly where a reel's energy lives and exactly what a casual viewer feels without being able to name.

What does the result look like?

From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 48.76 seconds, AAC audio. Vertical, forty-nine seconds, built to the reference's structure and pacing with a generated presenter and its own subject matter.

Forty-nine seconds matches the reference closely. Vizard Agent holds to the reference's length by default because pacing and duration are the same decision, and a structure stretched to twice its length stops being that structure.

The presenter is generated from a character sheet built before any scene, which is what keeps the same person in every shot and makes a second episode possible without starting the casting again.

When does this not work well?

Structure is fair to learn from and the material is not. Vizard Agent measures the reference and builds something new against those measurements, and the line between studying a reel and copying it is one you have to hold.

How do you fix a result that came back wrong?

Say which property is off. Vizard Agent keeps the reference's shot map, the pacing analysis, the generated character sheet and every music candidate it produced, so changing the tone or the speed is a re-cut against measurements it already took rather than a fresh study of the reel.

How does Vizard Agent compare to doing it yourself?

By hand this is watching the reel twenty times with a stopwatch, writing down every cut and every text pop, and then discovering your subject does not fit into the same beats because you never measured how long each one actually was.

By hand Vizard Agent
Understanding the reference Watch it repeatedly Shot map, transcript, overlay analysis
The pacing Copy by feel Measured as a property
A consistent presenter Film yourself, or skip it Character sheet generated once
Trying a different tone Rebuild the edit Re-cut against the same structure

Common questions

Does it copy the reel? No. Vizard Agent measures its structure and builds your subject into that shape with its own footage and audio.

Can it use the same music? No. The reference's track is licensed to them; Vizard Agent sources yours separately.

What if the reel is in another language? Fine. Vizard Agent transcribes it to understand the structure and builds yours in your language.

Can I be the presenter? Yes. Upload footage of yourself and Vizard Agent fits it to the structure instead of generating a character.

Will my version be the same length? By default yes, because length and pacing are the same decision. Say if you want it shorter.

Can I do a whole series this way? Yes. Keep them in one project and Vizard Agent carries the character and the styling across every episode.