How to Stylize Music Videos with AI

‍

Music videos are the game-changer in music.

‍

It helps you stand out but only if your music video illustrates the DNA of the artist and its music.

‍

That's why AI style transfer empowers everybody - from the freelancers creating music videos for artists to the 3D artists - to effortlessly transform their videos into visually stunning artworks.

‍

In this tutorial breakdown, we explore the step-by-step process of applying AI style transfer to a music video using Mago’s latest model, including tips on prompt design, ControlNet net settings, and render optimizations to unlock the potential of AI-powered video stylization.

‍

Getting Started with Mago

‍

To begin, simply create a new project in Mago and drag your source video into the timeline. In seconds, the video appears ready to be stylized.

‍

For this tutorial, we’re using Mago V3. V3 is our most stable and advanced model yet. If you prefer, V2.5 is also available, though with less consistent stylization results.

‍

Mago is in closed beta, join the Mago waitlist

‍

Crafting the Style with Image-to-Image Workflow

‍

Mago integrates an intuitive image-to-image workflow that supports both:

• Kontext workflow: Generate purely from a text prompt using the Kontext model.
• Mago workflow: Upload a style reference image, leverage ControlNets and prompts to create your image.

‍

In this tutorial, we use the Mago workflow to guide the stylization with a reference image. After uploading the image, a simple yet effective prompt like “colorful drawing style” provides creative direction.

‍

Fine-Tuning the Style of your Reference Image

‍

Customize your output using these advanced settings:

• Image Weight: Increase for stronger adherence to your style reference
• CFG Scale: Lower slightly to reduce prompt rigidity
• ControlNets: Guide the model’s interpretation (e.g., depth map, pose, etc.)

‍

Once set, click Generate.

‍

Multiple frames can be queued simultaneously, enabling fast iteration until you achieve the desired look.

‍

Pro Tip: You can choose the input frame you want to use at the top left of the image-to-image window.

‍

‍

Stylizing the Video with AI

‍

Once you have your reference image, you can start to stylize your videos. The setting panel is divided in three parts : Basic Video Tab, Advanced Video Settings Tab, ControlNets Tab.

‍

Basic Video Settings

‍

After selecting your preferred image frame, apply it as the style reference for the full video. You can:

• Use Mago’s automatic video and image descriptions to assist prompt generation
• Keep or adjust the negative prompt (we kept it default here)

‍

Pro Tip: Adjust the negative prompt only if you see something in your first stylized render that you do not want to be in the video.

‍

Advanced Video Settings

‍

Under the Advanced tab, we recommend:

• Model: Mago V3 for higher fidelity
• Steps: Increase slightly for sharper output
• Interpolation: Optional; useful for faster renders
• Color Match & Context Strength: Boost for stylistic coherence

‍

ControlNet settings

‍

• Control Net: With Mago V3, you can choose from four types; we used Depth
• Face Mesh: Enable for better lip sync when working with close-up facial footage
• Resolution: Bump it up for high-quality output

Run a 24-frame test render first to preview the result side-by-side with your input video (Control + Left Click for comparison).

‍

Pro Tip : Each ControlNet will be better for a certain type of video you want to get. Here is the list:

• Depth: For preserving spatial relationships and 3D volume

• Pose: For dynamic motion and character performance

• Normal Map (beta): For detailed textures and surface features

• Soft Edge (beta): For clear outlines and stylized contours

‍

Iterate and Expand: Building the Full Sequence

‍

Iteration is key in your optimized creative process. Mago follows this flow:

• Duplicate tracks using the shortcut Shift + D
• Each duplicated track inherits settings from the previous one
• Modify prompts, settings, or ControlNets incrementally for improved results

For longer renders, you can now generate up to 1,800 frames in Mago. In our case, we expanded to 100 frames and launched the final render directly.

‍

‍

Using AI Tools to Restyle your Music Videos

‍

This walkthrough demonstrates how AI style transfer with Mago turns your music video footage into stylized visuals, ready for music videos, animation projects or artistic experimentation.

Mago is designed for creative artists and video creators. With all its features while giving you creative control, you have all the tools to create stunning video transformations: no code, no messy node or compositing required.

‍

Ready to Try It?

‍

Start your next AI video experiment with Mago and see how far your creativity can go.

‍

Join the Mago waitlist

‍

‍