I Turned My Song Into a Music Video Using AI (One More Shot)

Added: 2026-06-03

[00:00:02]Making a music video used to be insane.

[00:00:08]You needed a crew, a location, a month of preparation, and a budget most artists don't have.

[00:00:19]Not anymore.

[00:00:22]Everything you just saw was made in here.

[00:00:26]This is one more shot.

[00:00:30]Let me show you how it works.

[00:00:35]In 2026, 87% of musicians are using AI somewhere in their workflow, and promotion is where they want it most.

[00:00:42]The game is changing fast, and we want you in the right side of it. Today, I want to show you how you can create a music video using our app. If you want to try it for yourself, the link for one more shot is in the description. Let's get right into it. There are two main modes inside and the one you pick depends on how much of the work you actually want to do yourself. Today I'll show you both of the options. So feel free to skip to the part you're most interested in. The first is auto. Auto is the fastest way to get from a song to a finished music video, letting the AI do all of the heavy lifting. Four steps. That's it. Step one, the type of video. We can see there are two options inside. performance and lyrics. Performance is used to create the classic video clip that we all know.

[00:01:31]Lyrics is an animated video built around the words of the track. Now, we'll concentrate on the performance video.

[00:01:38]Step two, the song. Here you can upload your song. It can be a song you've made or something you've generated. Just drag the track inside the page and that's it.

[00:01:49]For me, I'll use a song I've made before using Sunno. Once uploaded, you have the option to trim the song. If this is your first time trying out this mode, we recommend to try it on a bunch of seconds before doing the full length song. The minimum is 30 seconds, so I'll do that. Once you've selected the part that you need, we'll move to step three, the singer. This is who the video gets built around, so the choice matters more than it looks. The app will use the space across all the shots where it's needed. If you want to be the character inside the video, upload a good photo.

[00:02:23]The app works best with a clean, well-lit shot of your face, ideally a frontal portrait. Avoid heavy filters, group photos, or anything with strong shadows across the face. The clearer the reference, the more consistent the result. Otherwise, you can create an AI artist, a generated character that stays consistent across all your videos. We'll cover that in a dedicated tutorial. For now, I'll use one I've generated before.

[00:02:51]Then we move on. Step four, the customization.

[00:02:56]This is where you tell the app what the video is about. First, I'll write the song's name. Then, we move on to the description of the video. We suggest to keep the description short and simple.

[00:03:05]You have a max of 800 characters, so keep it tight. A short description of what happens in the video, a brief description of the outfit, where you want it to be located, and the overall mood. If you feel like your idea isn't good enough, let the AI suggest something based on the song vibe. Once the description is done, we choose a style for the video. I'll use realistic since I want a realistic looking video.

[00:03:31]Then, as your last step, you choose the format, either horizontal or vertical, and hit continue. From here, the process begins to be more automatic. You will be first presented with a preview of what the video will be about. some shots of the singer, the color palette, and a brief description of what the video will be about. There's something that doesn't convince you, you can always come back and adjust to your preference. For me, it already looks amazing. So, the last thing remaining to do is to press start generation. From here, our AI will start creating the video. Keep in mind that even if AI is super fast, this process can take up to an hour. The video will land in your dashboard when it's ready.

[00:04:16]So, in the meantime, let's look at what happens when you actually want to direct the thing yourself. Let's have a look at advanced.

[00:04:26]Advanced is the opposite of auto. You create the video shot by shot and takes more work, but it gives you all the control you need. We start from creating a new project. Give it a title, pick an aspect ratio, and write a small description.

[00:04:44]This is the editor page. At first, it can be overwhelming to understand everything, but trust me, it's very simple. We have two main spaces. On the left, there's the generation space where all your generations will be previewed.

[00:04:59]On top of it, you can read frames and shots. Frames are still images.

[00:05:04]Meanwhile, shots are videos animated from those frames. You always build the frames first and then animate them.

[00:05:11]Think of it the way a real production works. You storyboard before you shoot.

[00:05:16]On the right, there is the timeline where we can sync everything to the music. When creating a project, you will need to upload your song on this side.

[00:05:25]So, let's start from that. Once it's finished uploading, the page will change in a timeline like preview. The timeline reads the audio and gives you something to build against. To start generating, write a prompt and pick a model. Each model has a different cost, quality, and look. Nano Banana Pro gives you premium detail. One is cheaper and faster. You pick based on what the shot actually needs. Once you start to use the app, you'll understand which one to use for every situation. To write the prompts, we suggest to use your preferred AI of choice. This way, you can concentrate on the creative part and let the AI do the technical stuff. The first thing I ask for is a prompt to generate a change of the outfit of my character. For example, I'll send Claude this generate a Nano Banana Prompt to change the character clothes into a long flowing white dress.

[00:06:18]Keep all the rest identical. The image will be attached as a reference. I copy and paste the output into the prompt box inside the app. As said, we will use the image of my character as reference by uploading it here. Then in the prompt, I'll tag it. This way, I'm sure that the models have all the context to generate what I'm looking for. This looks really good. Now that I have all the reference images that I need, I can start working on generating all the first frames I need. Since this song starts with her singing, I want a closeup shot of her that sings. So, in Claude, I write what I want. Generate a prompt for Nano Banana Pro. I need a close-up shot of the character. She has a disheveled look, messy hair, smudged makeup. She's inside a dead forest at night. The lighting of the image should be like if she had a flash pointed at her face. The background is in complete darkness. Once I have the prompt like before, I paste it inside the text prompt box. Use her Polaroid as a reference and generate some shots. Usually, I'll create four so that if all of them come out bad, it means that I have to regenerate the prompt. I think I really like this one, so I put it on my favorites and I drag it into the timeline. This way, I can already imagine how the shot could be.

[00:07:27]Usually I'll create the first three to four frames and then animate. For the sake of the video, I'll animate directly this. To animate the image, I can click animate and the app shows me the shots page. It will be automatically used as the first frame of the video.

[00:07:41]Optionally, you can use an end frame to define where the motion lands, but it depends on the video model you use. In the text prompt, as similar as before, you have to describe what you want to happen into the shot. Camera motion, simple action. Usually the video models understand the general composition and animate it as natural as possible. One or two lines at max. Set how long you want the clip to be and generate.

[00:08:05]Once you are satisfied with the shot, you can just drag it into the timeline.

[00:08:09]Adjust the duration inside the timeline so that it matches your preference. By default, every cut is synced to the song, but of course you can modify it.

[00:08:18]Once you've inserted the clip, you can adjust which part of the clip to take.

[00:08:21]So, for example, here I'll try to match it when she starts singing. Then, since I want her to sing along with the song, I'll hit the lips sync button. This way, when I export the video, this shot will be lip-synced. Remember to toggle it on when you want this to happen. Try different models. See which one works best for you. It's a learning curve, but the more you create, the more you'll be fast and precise. Once the timeline tells the story you want, on the top right corner, there's the export button.

[00:08:47]The app will start to assemble all the clips, lip-s sync the ones you selected, and give you the final result at the end.

[00:09:11]That's one more shot. Two modes, one song, the video you actually wanted.

[00:09:16]We're at the start of something big.

[00:09:18]Music videos used to belong to the few who could afford them. Now they belong to anyone with a track and an idea. And this is just the beginning. New models, new features, new ways to create, all coming in the next few months. If this video helped you, hit like, subscribe, and drop your first one more shot creation in the comments. We'll feature the best ones on our socials. See you in the next one.