<aside> <img src="/icons/help-alternate_green.svg" alt="/icons/help-alternate_green.svg" width="40px" />
</aside>
When using Midjourney video, you don’t start with a prompt. You start with a frame, a still image, called a Starting Frame. We normally speak strictly about prompting here, but we’ll pause a minute to talk about the Starting Frame because it influences what’s possible to prompt.
<aside> 💡
You don’t start with a prompt. You start with a still image called a Starting Frame, which influences what’s possible to prompt. So choose a good Starting Frame.
</aside>
Here are steps to get to the video prompting process:
1. Choose your Starting Frame.
Sure, the [Animate] button is available on the [Create] tab for your own images. But as prompters, we’ll rarely use it, because its function is to be fun and easy and not require prompting at all. If you use it, Midjourney will make a guess about the appropriate motion for the image and give you a video automagically.
As prompters, however, we’re most interested in the choices available when we click on an image to view it by itself. Then we see all the available Animate Image buttons. And sure, you can animate automagically from here too with the “Auto” and “Loop” buttons, but as prompters we’re most interested in “Animate Manually” mode. We’re definitely here for Manual mode.

But first, a few words about selecting your Starting Frame:
<aside> 💡
You may want to avoid making videos from upscaled images. At the start of every video render there’s a downscaling process that takes the starting image to a resolution appropriate for 480p video. The more we have to downscale your starting frame, the more risk of glitch. Starting with the lower resolution images may help reduce downscaling artifacts. Maybe.
</aside>
2. Choose Your Animation Mode