Text to Video AI: Type a Prompt, Get a Clip

Write one sentence, get a 5-second portrait video back in a few minutes. Sisif runs an open-weights video model, charges $1 a clip, and never stamps a watermark on the export.

Text to video is the simplest way to make a clip: you describe the scene, the model renders it. Sisif keeps the loop short. The prompt box takes plain English, the default clip is 5 seconds in 540x960 portrait, and a 720x1280 render costs 30 credits. There is no editor to learn and nothing to install. Every clip is $1, and a $10/month plan covers 8 clips a month when you need volume.

A short prompt is one 5 s clip at $1. A timeline is a story, $1 per 5 s, priced before you pay.

See the specs

$1 a clipNo watermarkNo subscriptionCredits never expire

Sisif · terms
sisif.ai · rendered on our own GPUs
Watermarknone, ever
Credit expirynever
Failed rendercredits returned
Commercial useincluded
Subscriptionnone required
Modelopen weights, our GPUs
Agentspay per clip over the API
1 × clip, 5 s, 540x960$1
Total$1
Same terms on clips, and on every monthly plan.

What you get

Feature Sisif
Price $1 a clip, no subscription
Watermark None, on every plan
Cost of a 5-second clip 12 credits at 540x960, 30 at 720x1280
Model Open weights
Export MP4, portrait
Aspect ratio 9:16 portrait (540x960 or 720x1280)

How it works

01

Describe the clip

Write what should happen in 5 seconds: the subject, the style, the camera move. The prompt stays attached to the render, so you can see what made each clip.

02

We render it

An open-weights model on GPUs we operate turns the prompt into a 540x960 portrait clip, most of the time in 2-5 minutes. A failed render returns the credits.

03

Download and publish

One MP4, no watermark, yours to publish anywhere, commercial use included. Pay $1 for the clip; nothing recurring.

Prompt tips

  1. 01 Name the subject, the action and the setting in one sentence: 'a red kite over a foggy beach at sunrise'.
  2. 02 Add a camera move at the end ('slow pan left', 'handheld') to get motion instead of a still shot.
  3. 03 Keep it to 5 seconds. Longer clips cost more credits and the model holds a scene best at that length.

Frequently asked questions

How much does the Text to Video AI cost?

Every clip is $1. One payment buys one 5-second portrait render, with no subscription. Monthly plans start at $10/month when you want volume.

Does Sisif put a watermark on a text-to-video clip?

No. Every export is a clean MP4 with no watermark or logo, on a single clip and on a monthly plan alike.

How long does a text-to-video clip take to generate?

A few minutes. You type the prompt, Sisif renders a 5-second portrait clip with an open-weights video model, and you download the MP4 from your library.

Explore more

Recent renders, with the prompt that made each one 5 s · 540x960 portrait
01Three girlfriends clink flutes on a hidden sandbar; swimsuit-model poses.
02Epic overhead shot of a desert glamping site in Morocco: starlit sky, lantern-lit tents, travelers sharing stories around a crackling fire, Milky Way time-lapse spiraling above, soft acoustic outro.
03A barefoot couple wrapped in a linen sheet, standing on a balcony overlooking turquoise waves at sunrise
04“Cabana brunch playlist” – Fresh fruit platters, draped white curtains fluttering, Bluetooth speaker playing mellow R&B, and toes idly tracing patterns in the sand.
05“Champagne Sandbar Soirée” – Three girlfriends clink flutes on a hidden sandbar; one splashes sparkling water at the camera while the others strike carefree, swimsuit-model poses.
06beach, sexy, holiday, honeymoon, chill

Ready to make a clip?

Every clip is $1. No watermark, no subscription, and the credits you buy never expire.

Render a clip See pricing
Render a clip 5 s · 540x960 · no watermark · nothing expires