claudeVIDEO
40-second cockpit POV demo has Opus 5.5 writing motion graphics code
A multi-model pipeline: GPT Images generated the visuals, Grok handled motion, Opus 5.5 wrote code to overlay the dashboard UI, and Astra provided the voice.
Ming-Cheng Ho
@MemorysaverMFAArchived 2026.09.28Originally posted 2026.09.27
Ming-Cheng Ho @MemorysaverMFAOriginally posted 2026.09.27
用 Opus 5.5 做了一支 40 秒座艙 POV demo，多個模型接力完成，不是一次 text-to-video。 流程：GPT Images 出畫面 → Grok 做動態 → Opus 5.5 寫 motion graphics 的 code → Astra 配聲音。 第一人稱是刻意選的：鏡頭鎖在駕駛座，儀表 UI 交給 Opus 用程式疊。 https://youtube.com/shorts/VJlhAPPbves
x.com/memorysavermfa/status/2103940439397294235Archived 2026.09.28 10:36