How was INMU KING created?

This page contains affiliate links.

The creator Yajuman has confirmed using Suno and the production cost. His public posts do not identify the video-generation model. We first cover what he disclosed, then show how to make a similar MV yourself.

What you can check about the production of INMU KING

INMU KING is a song and music video released by Yajuman on August 10, 2026. The name YAJU-MC is used in the title of the work. On the internet, there are many questions like "What kind of AI was used to create it?", but the referenced post does not mention the name of the video generation model or the detailed production process.

However, Mr. Yajuman posted about the production cost saying, ``The video part is about 40,000 yen, and if you include Suno etc., it's just under 50,000 yen.'' At least from this we can confirm that Suno was used to produce the music and that the video part cost money. It is not possible to determine the number of times generated or the breakdown of costs from this post.

The number of times the video will be generated cannot be determined based on the length of the video. Since the footage that was not used and the number of times it was remade are not disclosed, it cannot be used to estimate that ``the same MV can be made for this amount of money.''

Steps to prepare songs, images of people, and videos

1. Write out where to switch the video in the lyrics

Enter the song into the editing software and mark the beginning of the lyrics where you want to change the video. For example, for the ``Iced Tea'' part, we decide on the required footage one by one, such as a close-up of a glass, for the ``King'' part, a person sitting on a throne, and for the chorus, a full body dance. The following is an example of a cut list when making your own MV, not the original production record.

At this stage, you can just arrange still images on top of the song. Make sure that the pictures don't change in the middle of the lyrics, or that the same composition doesn't continue too long before converting it into a video.

2. If you wrote your own song, use Suno's Custom to add lyrics.

[Intro]
さあ 王のおでましだ

[Verse]
アイスティー片手に 今日も待機中
再生ボタンで 始まる戴冠式

[Chorus]
お前 はしゃぎすぎ
朝まで 踊りすぎ
王様 まだまだ 終われない

[Outro]
王のおかえりだ

1990s Japanese pop anthem, energetic male lead vocal, grand choir in the chorus, bright synthesizers, driving drums, dramatic build-up, catchy repeated hook, 140 BPM

Eurobeat, fast four-on-the-floor beat, bright synth brass, punchy bass, energetic male vocal, call-and-response chorus, short instrumental breaks, 155 BPM

Hyperpop, distorted synth bass, glossy digital leads, chopped vocal accents, energetic Japanese vocals, explosive call-and-response chorus, playful mood, 160 BPM

3. Prepare a full-body image and a close-up of the face of the same person

For dance images, choose one that includes head to toe, and the arms can be distinguished from the body and background. For lip syncing, close up of the face close to the front. If it's a throne scene, include not only the person but also the throne and the room in the image.

First, decide on the face, hairstyle, and color of the clothes, then arrange the images for each scene. If you notice that your face has changed after the video is made, you will have to recreate the scene. If the hand or glass is already broken, please replace it at this stage.

4. Specify "what kind of movement to add" to the image

Insert images of scenes into functions such as Runway that create videos from images. For a cut of a throne, first specify one movement, such as ``The person remains seated, and the camera approaches slowly.'' If you ask them to walk, turn around, dance, and turn the camera all at once, it will be difficult to see which command caused them to break down.

Runway's Official Guide also guides you on how to describe people and camera movements, rather than repeating descriptions of clothes and backgrounds in images. Look at not only the beginning and end of the generated result, but also the hands, face, and background in the middle, and decide the range to use.

The camera slowly pushes in toward the seated character. The character remains seated and looks straight ahead. Continuous single shot, steady lighting.

The character slowly turns their head toward the camera, then holds eye contact. Locked camera, continuous single shot.

The camera slowly pans right across the glass of iced tea on the table. The glass stays still. Continuous single shot, steady lighting.

5. Give the dancing to Viggle and the lip sync to HeyGen

If you want to transfer a fixed choreography, use Viggle's Mix. Add a dance video that can be used in Add Motion, a full body image in Add Image, and generate it with MIX. Avoid videos where the person is cut off from the screen, and even include scenes where the person's arm crosses the face.

For a singing face close-up, add the face image and the relevant audio to HeyGen's Photo to Video. Test the mouth movement on one phrase before submitting the whole song.

The items you press on the screen and how to pass materials are summarized in How to make an INMU KING collage video.

6. Arrange the video on top of the song and match the lyrics and movements

Load the song and video into CapCut or Premiere, and place the song on the lower audio track and the video on the upper video track. Align the beginning of the video with the first lyrics mark you made, and cut off the excess before and after. If the generated video has unnecessary sound, the volume will be lowered for that video only.

If the dance starts late, move the video position. Align the lip sync to the same position as the audio passed to the generation, and avoid cutting out the silence at the beginning of only one side. Finally, write it out as a single song, play the beginning, chorus, and end of the song to make sure there are no gaps in the sound or black spaces.

The original video generation AI cannot be understood from the published information alone.

It's impossible to determine the tools used just by looking at the finished video: HeyGen because the face sings, Viggle because the dance is natural. Suno and production costs, which can be confirmed from the author's posts, are treated as original information. The above steps are an example for another producer to create a similar MV.

Try images, videos, and audio during part of the chorus

At first, only edit a short part of the chorus. By arranging cuts of full-body dances, singing faces, and props like glasses to the song, you can see which images need to be recreated. For areas where the close-up of the face does not suit, we will change the length to dance, etc., using the materials we can use to increase the length.

Read next article

What the creator has confirmed—and what remains unknown

INMU KING was posted by Yajuman on August 10, 2026. In a September post, the creator estimated roughly ¥40,000 for the video and under ¥50,000 including Suno and related costs. This is the creator's description, not a verified itemized invoice.

The statement does not identify the video model, the number of generations, prompts, editing tools or total production hours. Avoid confidently attributing a specific workflow or assuming the same budget would reproduce the original MV.

Consult the creator's own statements for confirmed details and treat look-alike production tutorials separately. YAJU&U is a different work by a different creator.

• Creator's production-cost statement ↗
• Compare YAJU&U and INMU KING →
• Find the original audio →

Reference materials

• Adobe: Compatible range of generation extension
• Yajuman's post about production costs (Yahoo! real-time search)
• Suno: What rights do I have with a paid subscription?
• Runway: Creating with Gen-4.5
• Runway: Performance Capture with Act-Two
• Adobe Firefly: generating video with partner models
• ElevenLabs: Speech generation from text
• Viggle: Animation functions such as Mix
• HeyGen: Official guide for Avatar IV Photo to Video