claudeVIDEO

Opus 5.5 builds 3D sets and gpt-image-2 renders camera angles for a multi-shot video

The author breaks down the workflow: Opus 5.5 builds a 3D set, gpt-image-2 generates an image for each camera angle, and after lip sync the shots are cut according to what is being said.

あいびぃ

@ivy432hzArchived 27 days agoOriginally posted 2026.09.30

Opus 5.5 builds 3D sets and gpt-image-2 renders camera angles for a multi-shot video (image 1 of 2)

Opus 5.5 builds 3D sets and gpt-image-2 renders camera angles for a multi-shot video (image 2 of 2)

あいびぃ @ivy432hzOriginally posted 2026.09.30

Open original on X

これの仕組みはとても単純で、Opus 5.5 で 3D セットを作り各カメラの視点として gpt-image-2 で画像を生成、リップシンクをつけたり動画にしたりの素材達を組み合わせ、カメラを喋りに応じて切替えています

This post quotes another post: “Building an AITuber with a 3D stage generated by Opus 5.5”

x.com/ivy432hz/status/2105269408537235572Archived 2026.10.01 10:22