How to Automate Long-Form AI Video Creation in N8N Workflows

• Character and object consistency: Feed reference images so characters, products, and environments stay consistent from scene to scene—even over long runtimes.
• Multi-agent pipeline: Specialized sub-agents handle research, screenplay, character design, narration, scene generation, and editing—end to end.
• Prompt-first control: Use a single prompt to define style, pacing, aspect ratio, tone, and naming. Context images anchor visual identity.
• Built for automation: Use the N8N node, a JavaScript API, or connect via MCP to orchestrate video at scale.

• A reproducible N8N workflow that generates a long-form video from a detailed prompt
• Optional reference images for on-brand visuals
• Auto-wait for render completion or run asynchronously and poll for results
• Post-process steps to store, notify, and publish your video

• An N8N instance (self-hosted or cloud)
• A Scrptly account and API key: https://scrptly.com/
• Optional: Public URLs for reference images (product packshots, brand elements, character sheets)

1. In N8N, go to Settings → Community Nodes → Install New.
2. Search for n8n-nodes-scrptly and click Install.
3. The node will become available in your node palette.

1. Open Credentials in N8N.
2. Click New Credential → Select Scrptly API.
3. Paste your API key from your Scrptly account page.
4. Save and select these credentials in your Scrptly node.

• Trigger: Choose how you’ll kick things off (Schedule for recurring series, Webhook for on-demand, or Google Sheets/Notion for content calendars).
• Scrptly Node: Configure your video parameters:
• Post-processing: Add nodes to store the video (S3/Drive), notify (Slack/Email), or publish (YouTube/TikTok integrations).

Create a 8–12 minute educational, documentary-style video titled "{{title}}" aimed at {{audience}}. Tone: authoritative yet warm. Structure the film into 6–8 chapters with on-screen chapter cards and smooth transitions.



Visual style: cinematic, natural lighting, color grade slightly warm. Mix close-ups and wide establishing shots. Maintain consistent character visuals from references if provided.



Include:

- Narration: clear, well-paced voiceover with coherent arc.

- B-roll: dynamic scenes supporting key points (diagrams, gestures, environments).

- On-screen text: concise, legible captions and key stats.

- Call-to-action at the end tailored to {{ctaGoal}}.



Constraints and preferences:

- Aspect ratio: {{aspect}} (e.g., 16:9 landscape for YouTube)

- Pace: moderate; pause for important visuals.

- Add tasteful background music with dips under narration.

- If context images are provided, treat them as canonical for character/product consistency.



Chapters to cover:

1) Hook and promise

2) Background and context

3) Core concept #1 with visual example

4) Core concept #2 with visual example

5) Objections and misconceptions

6) Practical steps and checklist

7) Summary and next steps

8) Closing CTA

• Trigger (Schedule): Run daily at 03:00 UTC.
• HTTP Request (Data Fetch): Pull the next topic from your CMS or spreadsheet.
• Scrptly (Generate Video): Feed the composed prompt and context images.
• If (Result): Branch on success/failure.
• Storage: Upload video to S3 or Drive.
• YouTube: Publish with title/description pulled from the same row.
• Slack: Notify the team with the video URL and thumbnail.

• Wait For Completion = ON: The Scrptly node blocks until the final video is ready—simpler orchestration.
• Wait For Completion = OFF: The node returns a task ID immediately. Add a polling loop (e.g., HTTP node to Scrptly’s task endpoint) or a wait-and-check pattern. This is ideal for high-throughput pipelines.

• Start with the default 10,000 tokens.
• Increase for more complex, longer, or research-heavy prompts.
• Monitor typical usage over your first week, then set guardrails to control spend.

• Be explicit in your prompt about recurring characters, camera language (e.g., "return to the same interview setup"), and palettes.
• Use 2–6 reference images per anchor entity (product, environment, or character).
• Provide chapter outlines and desired durations per chapter in your prompt.
• Standardize aspect ratio and resolution across your series for faster publishing.
• Add accessibility: captions, high-contrast text, and alt-text for thumbnails.

import Scrptly from 'scrptly';



Scrptly.setApiSettings({

  apiKey: process.env.SCRPTLY_API_KEY, // keep this secret

});



// From here, invoke the AI Video-Agent or VDK methods

// within your service and coordinate with N8N via webhooks.

• Long render times: Reduce runtime targets, simplify visual complexity, or run asynchronously and notify on completion.
• Inconsistent visuals: Strengthen references and emphasize consistency rules in your prompt (shots, wardrobe, locations).
• Context images not applied: Ensure image URLs are public and stable; list the role of each image in your prompt (e.g., "Use Image 1 as the main host reference").
• Budget overages: Lower Approve Up To, cut chapter count, or reduce research/detail requirements.

• Weekly YouTube documentary series based on your editorial calendar
• Long-form product explainers with consistent hosts and brand environments
• Research-driven thought leadership with chaptered structure and key takeaways
• Training modules with standardized layouts, captions, and callouts

• Install the N8N node: n8n-nodes-scrptly
• Spin up your first workflow and generate a long-form video today
• Explore the developer platform and API to scale programmatically