How to use OpenAI’s Sora to create stunning AI-generated videos

Browse to the Sora website and sign in with your OpenAI account. Before you try creating your own video, see what others have done. Select the Featured category at the top left to view the available videos. Go back to the home page and do the same with the Top and Recent categories. With the Top category, you can also change the duration to Day, Week, Month, or All time.

To start creating your own video, select the All videos category under Library. You’ll now want to choose a specific preset, aspect ratio, resolution, duration, and number of variations.

Click the Preset button at the bottom. Here, you can select among Archival, Film Noir, Cardboard & Papercraft, Whimsical Stop Motion, and Balloon World. Next, click the Aspect ratio button to choose 16:9, 1:1, or 9:16. Click the next button to choose the resolution and then the one after that to set the duration. With the next button, you can choose the number of variations of a video you wish to see — one, two, or four.

Finally, hovering over the question mark icon will tell you how many credits you’ll use with the settings you chose.

After setting the different attributes, type a description of the video you want at the prompt. You can be as detailed and specific as you’d like. One trick here is to turn to ChatGPT to generate the prompt. At the ChatGPT website, ask the AI to create a prompt and then provide a brief description of what you wish to see in your video. If you like ChatGPT’s description, copy and paste it back at the prompt in Sora. When ready, click the up arrow to submit your prompt.

With your video open and playing in the lightbox, Sora offers several ways to edit it.

Click the Edit prompt button to refine the prompt. Click the Edit story button to tweak the description for each action used to create the video. Click Re-cut to trim or extend the video via its timeline. Click Remix to describe any elements that you want to add, remove, or replace in the video.

Click Blend to blend elements from this video with another. You can then choose the second video from your computer or your library. Select Loop to put any section of the video on a perpetual loop.

Based on the change you specified, Sora will create one or more new versions of the video. Return to the All videos section in your library to view them.

With a standard video, Sora creates all the action for you based on its interpretation of your prompt. But you can better direct the video through a storyboard. Just like the storyboard that a director might use for a film, this option lets you describe and map out each action in the video. You can set up as many as seven actions in your storyboard.

Also: Gemini can now watch YouTube for you – skip the video, get the highlights

To get started, check out some of the storyboards in other videos. Return to the Feature, Top, or Recent categories. Select a video you like and click the View story button. Look at the prompts for all the actions that comprise the video to see how it was described.

Return to your own library and click the Storyboard button at the bottom. Click the icons at the bottom to choose an aspect ratio, resolution, duration, number of variations, and preset.

At the caption card for the first action, type a description. You can also add an existing video from your PC or Sora library. Since this will be the start of your video, be sure to include all the necessary details to establish the setting, characters, and action you want to occur.

If you have trouble here, you can again ask ChatGPT to write the prompts for you. Depending on how many actions you want to incorporate, describe what you want to see for each one. ChatGPT will then create multiple prompts. Copy and paste the first prompt into the first caption card back at Sora.

Next, you’ll want to space out the actions so that one smoothly segues into the next. Placing them too close together could result in jumpy transitions, while placing them too far apart could yield unwanted details added by Sora.

You may have to experiment, but keep in mind the total duration of the video as you place each card in an ideal spot. On the timeline, just drag and drop each card to its new location.

When you’re done, click Create to generate the video.

Return to the main screen to see your new video. Hover your mouse cursor over it to watch it play. Move your mouse left and right to scrub back and forth through the video. Click the video to view it in the lightbox. You can then tweak the video by revising any of the prompts, re-cutting it, remixing it, blending it, or looping it.