1254 hand-tested AI prompts across image, video, and music — each with full context, variables, and a sample output. Content is authored in Traditional Chinese; the UI here is English.
A complete single-shot narrative, from standin…
先用 Nano Banana 生兩張關鍵影格，再交給 Veo 補間成運鏡影片的進階工作流。
2026 是「圖轉影」之年——用一張角色定裝圖當 reference，讓 Sora 2 / Veo 3.1 生的影片人物前後一致、不走樣。
A cinematic 30-second brand story: three shots, with optional voiceover.
Veo 3.1 has the best recognition of cinematic camera terms (dolly, pan, rack focus). This prompt uses the right vocabulary to land a Dune-grade opening shot — perfect for the start of an MV, short film, or documentary.