AI Video Generation Pipeline
Idea in — story, cinematic VEO prompt, generated video saved to Drive.
Overview
An n8n form workflow that turns a short idea into a full story, converts it into a cinematic video prompt, submits to a VEO 3 generation API, polls until complete, and uploads the finished clip to Google Drive.
Problem
Raw video models need long, visual prompts. Going from a one-line idea to a usable clip means multiple manual LLM steps and API polling.
Solution
Chained two LLM stages — story writer then video prompt designer — before HTTP submit to VEO 3, wait/poll loop for task completion, conditional retry path, and automatic upload of the MP4 to Drive with a descriptive filename.
Architecture
n8n form trigger, OpenAI chains for story and video prompt, HTTP Request to VEO createTask, Wait node, poll recordInfo, If node on status, download/upload via HTTP and Google Drive nodes.
Workflow
- 01 User submits a video idea via form
- 02 LLM writes a visual, scene-rich story
- 03 Second LLM converts story to cinematic video prompt
- 04 VEO API generates video with async polling
- 05 Finished file uploaded to Google Drive
Results
- · One form replaces multi-step prompt engineering
- · Stories structured for motion and scene clarity
- · Async generation handled inside the workflow
- · Output archived automatically in Drive