跳到正文
@PixVerse_· @pixverse_ · X·· 2026-05-25AI 评分53
AI 导读

PixVerse 官方分享了一套在 PixVerse 网页端生成 15 秒视频的工作流,分两步完成。第一步让 ChatGPT 生成 12 格分镜,标注每个镜头的动作、机位和精确时间戳,例如 Shot 01 为 0–1.25s、Shot 12 为 13.75–15s。

正文

Workflow below!Try it at Pixverse web

Step 1: Generate Storyboard Ask ChatGPT to create a 12-panel storyboard of the Corgi video. Include each shot’s action, camera angle, and exact timestamp (e.g., Shot 01: 0–1.25s, … Shot 12: 13.75–15s).

Step 2: Generate Video Segments Upload your storyboard image to PixVerse and choose a model; for this project, Seedance 2.0 was used.

Type a prompt like: Create a 15-second cinematic video based on this storyboard. Generate each of the 12 shots in order, keeping the dog, apartment layout, and lighting consistent across all shots. Include subtle SFX (keys, door click, paw taps, tail thump) and a soft piano BGM. The AI will automatically turn the storyboard panels into cut-by-cut video segments.

Tip: Clearly specify the duration of each individual shot in your instructions so the model can generate footage with accurate timing and maintain smooth pacing across the entire video.

来源:@PixVerse_ · x.com