Genspark AI Tutorial: How to Make AI Videos in Minutes

Full video transcript

How’s it going y’all? Darby here with GI University. In this video, I want to share with you a way that you can generate videos with GenSpark. Now, Genspark is an AI super agent that can do a whole lot of things. Can’t go into all of the details of what it can do in just one video, at least not right now. But I can show you how I’ve been using it to generate AI videos and how you can also use it inside of GenSpark.

Once you select the AI video feature and inside of this chat, which probably looks familiar if you use ChatGPT or other chat-based large language model interfaces, you can select a mixture of agents. That means you can select Gemini V3. Of course, there may be some limitations to going through GenSpark versus going directly to any one of these, but these are also all going to have their own costs to each of them. And so, you can tap into that through GenSpark’s interface.

You can also do image to video. You can do horizontal, YouTube style, 3 to 5 seconds, 5 to 10 seconds, and you can prompt it to tell it what you want to do. So, I’m going to start with the demonstration for the Gemini V3. And again, your use case will probably vary depending on what you’re doing — whether you’re creating video sales letters or presentations, social media clips, or you’re looking for different forms of B-roll that you can incorporate into the videos and the demos that you’re doing for other types of video.

All you need to do is start with a prompt instruction that you want the AI to create. I’m going to give it a prompt here so you can see how it works. I want you to create an image of an iguana hanging out on a hammock that’s overlooking a lake with boats and things like that happening on the lake. It’s a sunny day. There’s nice trees around for shade, and the iguana is just hanging out swinging back and forth and he’s playing some drums.

So with that, I’m just going to go and give it this prompt. Noticing right here, public creative recreation is okay to share with the community, inspire others, or you can keep it private. In this case, I’m going to go ahead and select private for that. So, just keep an eye out on how your settings are set up inside of your account whenever you’re using this.

And this is what it gave back to me. This is what it means when it’s saying autoprompt. You’re looking to create a captivating video scene featuring an iguana leisurely hanging in a hammock, yada yada yada. You can see what it’s saying here. And then it is creating these autoprompts. And so as we go down here, we can see this autoprompt that basically took the idea and the content that I gave it. And this is interesting because it’s giving language that you might use if you were a director with more specific shots around how you want the scene to be captured. And so the more well-versed you are in that sort of language, I imagine you’ll be able to explore different levels of video generation based on the detail you provide. It’s like: shot in a wide-angle view to capture the entire serene lake scene with an emphasis on natural ambient light. So again, the more context you could provide for exactly what you wanted inside of this particular scene, the better outputs you’re going to get.

And so with that in mind, there’s also ways that we can incorporate things like the image to video and have it actually take an image that we’ve created and turn that into a video. But we have an example right here. Right. So, we can see here from that example, this generation happened. I actually ran through a few more, so I’ll just save you guys the generation of all of this content. And this next one, I uploaded another image to give it some more context. So, I uploaded this image. It gave me back another reflection. And then it output a much more aligned video actually based on that reference image. So, it can do reference images.

And then I asked it to do it again but with a different model, the Mini Max, which you can see when you look through the different models, the Max right here. And it created a much more realistic version. But I will say that on this generation, I didn’t resubmit that image that I shared with you just now that I gave to that last generation. And so it did this one — it plays this one, a much more realistic iguana playing bongos on a lake. A lot more people in the background, a lot more detailed. Looks like it’s having a fun time.

And then I resubmitted the image. I asked it to reference this image again. But I wanted to try this time with Kling as another model. If you have used different AI tools before for video generation, you may have seen Kling out there as well. And it did create an image. You can see the background — if you notice, it is changing slightly. There are images moving inside of this 10-second loop.

And yeah, so I would encourage you to try out the different models inside of GenSpark. Give it a go, see which one works best for you and what you’re looking to create and produce.

For more content like this, like, subscribe, and share. See y’all on the next video.