Vizard Agent

How to make a Reel where every sentence gets its own matching shot

Last updated 2026-08-17 · 6 min read

Tell Vizard Agent the topic and ask for the visuals to match each sentence. Vizard Agent checks the facts it is about to state, records the narration, sources a shot for every line, builds animated headline cards, and re-renders the cut when the zooms or the footage do not carry the sentence they sit under.

What is the short version?

A Reel about a topic works when the picture agrees with the sentence being spoken. That means one shot per line rather than a montage under a voiceover, and it is the difference between a video people follow and one they scroll past.

  1. Go to Vizard Agent and say the topic and what it should cover.
  2. Ask for the visuals to match each sentence.
  3. Say the treatment: captions, subtle zooms, smooth transitions, modern music.

What do you need before you start?

The topic and the treatment. Vizard Agent researches, writes, narrates and sources, and the useful thing you add is the finish — because "match visuals to each sentence" is a structural instruction that changes how the whole reel is built.

What do you type into Vizard Agent?

Say the topic and ask for the sentence-by-sentence structure explicitly. Vizard Agent handles the research, the script and the sourcing on its own, and that one structural phrase changes the finished reel far more than any adjective about style or pacing would.

Prompt

Create a vertical video about [topic]. Match visuals to each sentence. Add captions, smooth transitions, subtle zooms and modern background music. Around [50] seconds.

Variants worth knowing:

What does Vizard Agent actually do?

Here is the order Vizard Agent worked in on a real topic reel. It starts by checking what it is about to assert, and it spends the second half correcting things it did not like about its own first cut.

  1. Loads the playbook and checks the latest news on the topic.
  2. Double-checks the release date facts before writing anything.
  3. Checks the available tools and their prices, and reads the caption style presets.
  4. Records the narration and gathers stock shots in the same pass.
  5. Checks the narration's length and the stock candidates, then downloads the clips and music.
  6. Builds a contact sheet of the footage and looks at it.
  7. Times the narration and builds the animated headline graphics, then tests how to decode the transparent overlays before using them.
  8. Renders previews of the headline cards and checks they look right.
  9. Assembles the picture cut, reviews it, sources two more night-city shots, and re-renders with corrected zooms before generating the transition sound effects.

Step 2 is the one that matters most in a topic reel. A stated release date is a claim, and Vizard Agent verified it twice before putting a voice behind it.

What does the result look like?

From the run this page is written from, probed on the delivered file: 1080x1920, H.264, 30fps, 51.3 seconds, AAC audio. Vertical, fifty-one seconds, one shot per sentence with animated headline cards, captions, subtle zooms and transition sound effects.

Vizard Agent went back for two additional shots after reviewing its own picture cut, rather than accepting footage that did not carry the lines it sat under.

Vizard Agent keeps every sourced clip and the narration's word timings after delivery, so replacing a single shot later slots the new footage into exactly the same span of speech without disturbing anything around it.

This category was not separately measured, so treat the timing as a range rather than a promise: comparable work runs a median of 28 to 38 minutes end to end. Across all projects the median cost by tier is Flash 47, Pro 55, Max 242, Ultra 263 credits.

When does this not work well?

A topic reel states things out loud, which makes it the format where being wrong is most expensive. Vizard Agent grounds the script in what it finds and double-checks the facts it is about to assert, and news about a topic still moves faster than any video about it does.

How do you fix a result that came back wrong?

Name the sentence that is wrong. Vizard Agent keeps the narration, its word timings, every downloaded clip and the headline graphics as separate assets, so replacing a single shot or correcting one spoken line leaves the rest of the reel exactly as it was.

How does Vizard Agent compare to doing it yourself?

By hand this is research, a script, a recording, and then a separate stock search for every single sentence — a dozen searches — plus animated cards built in a design tool and a caption pass, all for one video about one topic.

By hand Vizard Agent
Checking the facts Read one article Verified before writing
A shot per sentence A search each Sourced and reviewed as a set
Headline graphics Build in a design tool Generated and preview-checked
A weak shot Ship it Replaced after reviewing the cut

Common questions

Does it research the topic? Yes, and it double-checks anything datable. You should still verify anything time-sensitive yourself.

Will every sentence get its own shot? Yes, if you ask for it. It is a structural instruction rather than a stylistic one.

Can it use my own footage? Yes, and it prefers it. Vizard Agent uses your material first and sources stock only for the lines it does not cover.

Can it add animated headline cards? Yes. Vizard Agent generates them and checks the transparent overlays render before using them.

How long should a topic reel be? Around fifty seconds for several beats; thirty if you want it tighter.

Will the captions match the narration? Yes. They come from word timings measured on the generated read rather than estimated afterwards.

Can it add transition sounds? Yes. Vizard Agent generates them and places them on the cuts.

Can I get a shorter version? Yes. Vizard Agent rewrites to fewer sentences rather than speeding the same read up.