Load an image and an audio file, add a prompt, and set the width and height to match the aspect ratio of your uploaded image. Check the workflow for the recommended resolution. Then you’ll have a panel with switches that lets you choose how long you want the video to be, from 3 to 34 seconds.

Generate.

It’s good for speaking, but you can’t do complex animation like walking and speaking at the same time. You can prompt things like hand movement and small gestures, but it works best for talking heads.

Keep the duration in mind based on your subscription. Some plans allow a maximum of 20 minutes, others 40 minutes, and so on. So if you try to generate 34 minutes and it takes around 50 minutes to complete, but your plan only allows 40 minutes, it will stop the workflow before it finishes. Alternatively, you can enable Save output for each video and Combine output, and then it will save each 3-second video segment.

For example for 6 second of video using 48gb vram video card it took around 8 minutes for 720x1280px, or you can do smaller like 480x832px to do it faster.

Don't forget to enable save output for the last video combine you use.

Learn ComfyUI for free from Pixaroma Youtube Channel https://www.youtube.com/watch?v=HkoRkNLWQzY&list=PL-pohOSaL8P-FhSw1Iwf0pBGzXdtv4DZC