3D, VR & spatial ·
How this music video was made with zero image AI
No image or video model was used anywhere in this music video. Claude Opus 5.5 got an MP3 and about 15 short notes, then wrote every frame as code: 9 chapters, around 50 shots, each cut snapped to the beat.
It analyzed the song, matched lyrics to the vocals, drew every character and set procedurally, then rendered test frames, looked at them and fixed its own mistakes. The final render was 4,698 frames at 1080p, painted in software over two hours.
Made by u/DigitalDaydreamers1. This write-up follows the original post; it is not an independent product test.
Keep up with what gets built.
Read the daily AI briefing, browse the tools, or sign up for briefing email updates.