See llms.txt for all machine-readable content.

Back to Templates

Create and review AI short videos from Google Sheets with OpenAI, fal.ai, and Creatomate

Created by

Created by: Andy || banandyy
Andy

Last update

Last update 8 days ago

Categories

Quick overview

Turn approved Google Sheets briefs into vertical AI videos with Seedance scenes, an instrumental soundtrack and timed captions. Save provider jobs between runs, inspect the finished MP4, and use a separate authenticated review workflow before optional publication through Blotato.

How it works

  1. Reads one approved, uniquely identified brief from a Google Sheets Queue and loads or creates its authoritative, versioned n8n Data Table record.
  2. Uses OpenAI structured output to create a bounded scene plan, then submits Seedance clips and an instrumental Stable Audio soundbed through fal.ai.
  3. Saves each request intent before HTTP and stores returned provider job IDs. One durable operation runs per manual or scheduled execution; subsequent runs poll or resume the existing jobs.
  4. Builds a Creatomate render with explicit scene timing and caption overlays. Downloads the completed MP4 and checks its container metadata, dimensions, duration, audio stream and SHA-256 bytes.
  5. Mirrors production status into Google Sheets while the Data Table remains the authoritative record. Provider media URLs are temporary, so the finished video should be downloaded promptly.
  6. The separate companion workflow downloads the actual MP4 review ZIP, records exact-version approval, and optionally submits one approved Blotato destination at a time. Approval binds media, text, account IDs, privacy and AI disclosures.
  7. Stops uncertain submissions for manual reconciliation instead of blindly repeating paid generation or publication requests.

Setup

  1. Run the credential-free local demo to inspect a synthetic MP4 and ZIP without external services. It is not a sample of live AI generation.
  2. Create the Queue sheet and native Data Table with the exact schemas in the overview documentation, or import the companion setup utility in the same n8n project. Copy the returned resource IDs into the factory configuration.
  3. Connect Google Sheets OAuth2, OpenAI API, fal.ai Header Auth and Creatomate Bearer Header Auth to the documented nodes. Set a supported structured-output model, current provider cost estimates and a per-job estimate limit before enabling generation.
  4. Add a unique brief with approved_for_generation=TRUE and production_status=ready. Run the worker repeatedly or deliberately activate its minute schedule after testing. Each execution advances one durable operation.
  5. For review, import the authenticated companion, set actual reviewer emails and matching state configuration. Keep publishing disabled until the video has been inspected, the intended Blotato account IDs and privacy settings are configured, and a live publication test is authorized.

Requirements

  • Designed for n8n 2.28.6 with native Data Tables and authenticated Forms. Production generation requires Google Sheets OAuth2, an OpenAI API account supporting strict structured output, fal.ai access to Seedance Pro Fast and Stable Audio, and Creatomate credits. Optional publication requires Blotato plus your own connected accounts. Model/API/rendering/hosting charges are separate from the template price. Tests used local synthetic media and mocked APIs; paid provider generation, actual social publication and n8n Cloud were not tested.

Customization

  • Edit Queue briefs, supplied brand facts, language, style, requested destinations, scene count and duration within the documented bounds. Configure the model, brand label, operator-maintained cost estimates, poll limits, reviewer allowlist and approval expiry. Configure destination account IDs, privacy and AI disclosures before approval. Changes to schemas, model endpoints, caption layout or validation require code review and retesting; changing an existing Queue row does not revise an already registered job.

Additional info

This is an independent implementation inspired by a public video-automation use case, not a copy or resale of another template. It creates instrumental audio and per-scene text overlays, not voiceover or speech-aligned subtitles. All exports are inactive; generation and publishing are off by default. Cost estimates are not billing caps and exclude OpenAI token costs. Provider media URLs expire. The package contains the companion review/setup workflows, English documentation, examples and dated test evidence.