One sentence → full episode

Text to video generator

Type a sentence. Get back a finished vertical episode with characters, spoken dialogue, sound and burned-in captions — not four seconds of silent footage that still needs a video built around it.

Story

Your one-line premise becomes a scene-by-scene script with a conflict and a payoff.

Cast

Original characters are designed and kept consistent from scene to scene.

Voices

Every character gets a distinct AI voice — dialogue is performed, not subtitled.

Captions

Word-timed captions are burned in, styled to stay readable with the sound off.

Export

One finished 9:16 episode for TikTok, Shorts and Reels. No editor, no export settings.

Text-to-video usually means text-to-shot

The category name is misleading. Almost every tool sold as text-to-video converts a prompt into a single silent shot of a few seconds. That is a genuinely hard technical achievement and it is also not a video — you still have to write the story, produce voices, order the scenes and add captions.

The gap between a shot and a video is the entire job. Frutti runs that job end to end: your sentence becomes a scene-by-scene script, an original cast, performed dialogue, animation, sound and captions, exported as one file.

Real outputs, not a promo reel

These are finished, self-hosted Frutti episodes — the same engine and pipeline this generator runs. Watch them end to end and judge the output before paying anything.

Fruit crime-story episode · 0:39

A multi-character fruit mystery built around a visible conflict, fast scene changes, and a clear final payoff.

Bone Lab · Dr. Marrow · 0:51

An absurd skeleton-scientist cast with spoken dialogue, scene changes, and burned-in captions — the same episode pipeline behind every Frutti render.

Locker-room fruit drama · 0:31

A finished vertical episode with a recurring fruit cast, spoken dialogue, motion, sound design, and burned-in captions.

Sentences that become episodes

Specificity is the whole input. A named character with a concrete problem produces a scene; a topic produces filler.

  • a mafia watermelon holds a family meeting about who left the fridge open
  • an owl runs a night-shift call centre for animals who cannot sleep
  • a skeleton news anchor reports live from inside a haunted vending machine
  • two onions argue about which of them makes people cry harder
  • a pomegranate lawyer defends a banana accused of slipping on purpose
  • a retired greyhound explains why he now delivers mail

The same text, next week, same cast

Because characters persist between renders, a second prompt continues the world rather than starting a new one. That is the difference between generating videos and building a channel.

Frequently asked questions

How long does a prompt need to be?

One sentence is enough, and longer is not better. What matters is that it names a character and a concrete conflict rather than a theme.

What exactly comes back?

A single finished 9:16 MP4 with performed dialogue, sound and burned-in captions — no separate audio file, no caption track to sync, no timeline.

Can I write in a language other than English?

Yes. The episode is written in the language of your prompt, and the voices and captions follow it.

Is writing free?

Yes — the full episode is written scene by scene for free without an account. Only rendering the video is paid.

Text to Video Generator — A Whole Episode, Not a Shot · frutti.ai