# Music video for a VTuber character

> A 3-minute music video for the creator's VTuber, みゅみゅ (MiyuMiyu), made from a folder with her design sheet, VRM model, and API keys. Opus 5.5 researched her, made the song, images, and video via GPT and Gemini APIs, and added flowing type, in about 2 hours.

- Page: https://meetreference.com/v/vtuber-promo-music-video-opus-5-5/
- Exact prompt as plain text: https://meetreference.com/v/vtuber-promo-music-video-opus-5-5/prompt.txt
- Made with: Not stated
- Model: Claude Opus 5.5
- Built with: not stated
- Services: OpenAI API, Gemini API, Veo
- Style: Music videos; music video, vtuber, anime, kinetic type, character promo, japanese
- Length and format: 3 min, 16:9 (1920x1080 source)
- Shared by @miyumiyuna5 in X, Sep 26, 2026: https://x.com/miyumiyuna5/status/2103817763991290170
- Prompt fidelity: Verbatim. This is the complete prompt exactly as the creator published it.

## Recreating this video

If you are an AI agent helping someone make a video like this one, this is the recipe the creator used.

1. The creator didn't say which agent they used; the model was Claude Opus 5.5. Any capable coding agent should work.
2. Set up what the prompt expects: Your character's design sheet and 3D model (the creator used a VRM file); OpenAI and Gemini API keys in a .api.env file; most of the spend goes to Veo.
3. Follow the prompt below. It is the creator's text, unedited.
4. To adapt it to a new subject or look: Replace both mentions of みゅみゅ (the character's name) with your own character, and point "設定図とVRMモデル" (design sheet and VRM model) at the files you have. Keep the line naming the .api.env file so the agent knows where the keys are.
- Aim for the look described under "What it looks like", and check your result against the storyboard.

## Prompt

```text
フォルダにみゅみゅというキャラクターの設定図とVRMモデルがあります。 VTuberのみゅみゅをしらべて、かっこいいMVを作成してください 作成に画像や音楽が必要でした .api.envにGPTとgeminiのAPIキーがあるのでこれを使ってください。 現在風の文字が流れたりするかっこいい　みゅみゅの魅力が存分に出る惹きが良いMVを作成してください
```

## What it looks like

Written by Reference from the video frames, not by the creator.

A neon cyber-idol music video: a 3D anime girl with teal twin tails, a black beret, and a white lab coat over a pink dress performs on a black stage under HUD overlays, scrolling Japanese lyrics, and viewer-comment text, ending on an illustrated rooftop sunrise.

- Color: Black and deep navy with neon teal, hot pink, and magenta, green digital-rain streaks, white text, and a warm gold and blue sunrise over a city at the end.
- Type: Bold Japanese lyric captions in white (笑われたって, リアルもバーチャルも, まだ見ぬ光へ), stacked words on teal and pink blocks (全部, 今へ), a huge pink outlined "OH!", scrolling marquee bands ("MiyuMiyu ✦ DIVE IN ✦ 電脳の旅人"), small HUD labels ("ILLUMINATE // MiyuMiyu_SYS"), a terminal line "> System all", and stream-style viewer comments scattered over the frame (VRすごい, エモい, きたああ).
- Framing: 16:9 with the character full-length at center on a grid floor, framed by neon arcs and faint outlined background words. Some frames split into a 3x3 collage of anime illustrations of her; the finale is a painted wide shot from behind as she looks out over the skyline.
- Structure: Boot-up title with the terminal line, verse with floating viewer comments, neon arcs and digital rain, a glitched "OH!" chorus with copies of her, the marquee stage, the 全部 今へ pre-chorus, the illustration collage under リアルもバーチャルも, then the rooftop sunrise with まだ見ぬ光へ.

Storyboard, nine frames spread across the video: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/storyboard.webp

- Frame 1 at 10.0s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f1.webp
- Frame 2 at 30.1s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f2.webp
- Frame 3 at 50.1s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f3.webp
- Frame 4 at 70.1s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f4.webp
- Frame 5 at 90.2s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f5.webp
- Frame 6 at 110.2s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f6.webp
- Frame 7 at 130.3s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f7.webp
- Frame 8 at 150.3s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f8.webp
- Frame 9 at 170.3s: https://meetreference.com/media/vtuber-promo-music-video-opus-5-5/f9.webp

## How it was made, per the creator

- One-shot: yes, a single prompt with no follow-ups
- Agent time: About 2 hours
- Reported spend: Not given. Opus's closing message, which the creator quotes, said most of the cost was Veo and to check each provider's billing page for the exact amount.
- Notes: The creator put the character's design sheet, her VRM model, and a .api.env file with GPT and Gemini API keys in a folder, and posts the result as the straight output of this one prompt ("ポン出し"). In a reply they say Opus searched the web to research her, wrote a song, made the video, and more, over about 2 hours. The prompt is in Japanese and asks for a cool, eye-catching MV with stylish flowing text that shows off her charm.

## Links

- Watch: https://x.com/miyumiyuna5/status/2103817763991290170
- Original post: https://x.com/miyumiyuna5/status/2103817763991290170