All articles
StudioAugust 20, 20265 min read

A video ad from one phrase: how the EYWA studio works

You type “make an ad for my dry cleaning” and get a finished file with sound. Here's what happens in between.

What happens after one phrase

Ordinary video models return a short silent fragment. After that you write the script, find a voice, pick music and stitch it together in an editor. EYWA takes that work on itself.

  • The director splits the ad into scenes and writes the script so the scenes cover the whole duration with no gaps.
  • Writes the voiceover text and performs it.
  • Picks and generates background music.
  • Adds sound effects where the scene calls for them.
  • Mixes audio with video and assembles the final file.

Prompt

«Make a 30-second Instagram ad for a carpet cleaning service»

Result

A finished vertical ad: product scenes, voiceover, music matched to the cut, and a closing frame with contacts.

Why the voice and music don't drift apart

When an ad is assembled from separately generated pieces, everything drifts between them: voice tone, loudness, the character of the music. EYWA generates the audio tracks separately from the video and mixes them across the whole ad.

So the voice sounds the same from the first second to the last, the music runs as one line, and you set the duration — not the limit of a single clip.

The ad can speak in your voice

Record a short sample of your speech — EYWA clones the voice and narrates the ad with it. This works precisely because the voiceover lives on its own track: it can be replaced without rebuilding the video.

If you don't need your own voice, pick a ready one — female or male, calm or upbeat.

What we need from you

At minimum, one phrase. A product photo helps: the studio will show your actual product on screen instead of a generic image. EYWA then asks a couple of clarifying questions, shows the cost in credits and builds the ad once you confirm.

Plainly about the models: text answers come from EYWA's own model on servers in Uzbekistan, while images, video and speech synthesis run on connected external models. Directing, scripting, mixing and assembly happen on our side.

Ask your first question

EYWA is easier to grasp in action. Open the chat and type what you need right now.