RIFT
Get Rift

How to make AI videos without learning to write prompts

Updated · Reviewed by The Rift team

Short answer

Skip the prompt tricks. In Rift you show Spark, Rift's video agent, a video you like or a photo, and say what to change in plain words, like 'put me in this' or 'same thing at night'. Spark works out the shots, camera and light, then shows a Draft you can check before you tap Make. Spark replies in the language you write in.

Show Spark a video or a photo and say what to change in plain words, like ‘put me in this’. Spark works out the shots, camera and light for you.

Show, then say

The clearest instruction is an example. Share the video into Rift and Spark already knows the moves, the camera, the timing and the place. Then you only say what’s different.

When you start from nothing

Say who, where and what happens, in one sentence. That’s it.

Check the Draft

Before you spend anything, Spark shows a Draft: its plan for the shot and the price on the Make button. If something’s off, say it plainly: “less dark”, “make it me on the right”, “shorter”.

Lines that work well

Tips

A good place to try it is putting yourself in a TikTok trend.

Questions people ask

Do I need to describe the camera and light?

No. Spark picks them. Say them only if you care, like "slow motion" or "at sunset".

What if Spark gets it wrong?

The Draft shows the plan before you spend anything. Tell Spark what to change and it updates the Draft.

Can I write in Spanish, Japanese or another language?

Yes. Spark replies in the language you write in.

Can I reuse a setup that worked?

Yes. Save the Draft as a format and run it again with a new photo.

Also in: Dansk · Deutsch · Español · Français · हिन्दी · Italiano · 日本語 · 한국어 · Bahasa Melayu · Nederlands · Norsk · Polski · Português (Brasil) · Svenska · Українська · Tiếng Việt