How to use Kling text-to-video and image-to-video

Kling is a state-of-the-art AI video generator. It takes a text or an image prompt to generate a short video clip. In this article, you will learn how to use Kling to create videos like the ones below.

What is Kling?

Kling is a video AI model developed by the Kuaishou AI Team. Upon its launch, Kling quickly became an internet sensation. It is widely regarded as a respectable rival to OpenAI’s Sora —except that you can use it today.

How to use Kling?

Kling can only be used as a web service. You cannot install it on your local computer.

Follow the steps below to start using Kling.

1. Go to KlingAI.com
2. Click Sign in for free credits in the upper right.
3. Complete the registration process.

You should get some free credits that are enough to generate a few videos.

Text-to-video

To generate an AI video from a text prompt:

1. Go to the video generation page.
2. Select Text to Video above the text input box.
3. Enter a prompt that describes the video. E.g.

a woman spy firing a handgun and running in a tesseract, dark scene.

3. Click Generate.

Image-to-video

To further control the video, you can supply an image as the initial frame. I recommend using an image generated with Flux or SDXL models.

1. Go to the video generation page.
2. Select Image to Video above the text input box.
3. Drop an image to the input canvas.
4. (Optional) Enter a prompt describing the video.
5. Click Generate.

After a while, you should get a video.

Tips for using Kling

Building a prompt for Kling AI video is not very different from making prompts for Stable Diffusion. But you should add some motion keywords.

Unlike AnimateDiff, Kling really uses the text prompt to condition the model for video creation. So, it is important to describe the video’s motion. The basic structure is

1. Subject – What’s the main character of the video. You should describe it in detail.
2. motion – What the subject is doing.
3. scene – What’s the surrounding.