How To Generate Long-Form AI Videos In N8N

• Try Scrptly: https://scrptly.com/
• Scrptly n8n community node: https://github.com/ybouane/n8n-nodes-scrptly

• End-to-end automation: Trigger from a webhook, schedule, form submit, or CRM event.
• Consistency at scale: Use context images so characters, products, and environments remain consistent across minutes-long videos.
• Developer-friendly: Scrptly offers an API, VDK (Video Development Kit), and MCP server; build flexible systems, not one-offs.
• Multi-channel ready: Output works for YouTube, TikTok, Instagram, ecommerce PDPs, and internal training.

• An n8n instance (self-hosted or cloud).
• A Scrptly account and API key: https://scrptly.com/
• Optional: Product/reference images you’ll pass as context images.

1. In n8n, go to Settings > Community Nodes > Install New.
2. Search for: n8n-nodes-scrptly
3. Click Install to add the node to your instance. Source: https://github.com/ybouane/n8n-nodes-scrptly

1. Open the Credentials tab in the n8n editor.
2. Create New Credential and choose Scrptly API.
3. Paste your Scrptly API key and save.

• Trigger: Schedule (e.g., weekly) or Webhook (e.g., after a CMS update).
• HTTP Request (optional): Fetch product images or a research brief.
• Scrptly Node: Generate the video.
• Storage/Publish: Upload to cloud storage, YouTube, or your CMS; share via Slack/Email.

• Prompt: The heart of your video. Describe the story, structure, pacing, scene beats, voiceover style, and visual aesthetics.
• Context Images: Optional but crucial for long-form consistency. Provide product photos, character references, or brand elements.
• Approve Up To: Set a max budget in tokens for longer videos.
• Wait For Completion: On (default) blocks until the video is ready; Off returns a task ID for async polling.

Title: Autumn Launch – Sustainable Leather Bag

Goal: Create a 4–6 minute product story that blends lifestyle footage, macro product shots, and customer testimonials.

Audience: Eco-conscious shoppers, 25–40.

Format: 16:9 landscape, YouTube-first; also produce a 60s short cut for social at the end.

Style: Warm cinematic, natural light, steady cam; subtle lofi soundtrack; calm, confident narrator.

Structure:

- Cold Open (00:00–00:20): Visual hook of city mornings + quick product tease.

- Segment 1 (00:20–01:30): The problem and how we solved it (materials, craftsmanship).

- Segment 2 (01:30–03:30): Real-life use cases (commute, coffee shop, weekend trip).

- Segment 3 (03:30–04:50): Testimonials (2 voices), emphasize durability and design.

- Closing (04:50–05:30): Brand values, call to action.

- Social Cut: 60s punchy highlight reel with subtitles.

Narration: Friendly, trustworthy, mid-tempo. Add captions.

Visual Consistency: Keep the same model and bag across all scenes.

• Use an HTTP Request node to fetch your product images or map images from a previous step (e.g., Airtable, Google Drive, Shopify).
• In the Scrptly node, pass these as Context Images. This ensures your character, product, and location appearances stay consistent across scenes.

• Turn off Wait For Completion to immediately get a task ID.
• Add a Loop or Polling routine (e.g., with a Wait node + HTTP Request if your flow needs custom checks) to periodically query status.
• On success, proceed to upload/publish; on failure, route to an error handler.

• Set a retry strategy: exponential backoff with a max retry count.
• Branch errors to a notification channel (Slack/Email) with context (task ID, prompt, timestamps).
• Log outcomes to a datastore (e.g., Google Sheets, Postgres) for auditing and iteration.

• Use scene beats: Spell out segments with durations. This helps pacing and narrative logic.
• Provide context images: Upload character angles, product macro shots, and environment references.
• Be explicit about style: lighting, color grading, camera motion, voice timbre.
• Set budget wisely: Increase Approve Up To for longer scripts to allow the agent to plan, render, and edit thoroughly.
• Reuse prompts as templates: Store them in n8n so your team can trigger consistent, on-brand videos.

1. Documentary Explainer (6–8 min)

• Objective: Explain a complex topic with clear structure, interviews, b-roll, and motion graphics.
• Prompt notes: "3 acts, chapter cards, lower-thirds for speakers, subtle ambient soundtrack, calm voiceover, captions on."

1. Faceless YouTube Automation (8–12 min)

• Objective: Produce recurring episodes without on-camera talent.
• Prompt notes: "Narration-led, stock-style b-roll aesthetics, consistent iconography, stylized infographics, chapter markers, end-screen CTA."

1. Educational Series (5–7 min each)

• Objective: Multi-episode course content with consistent characters and setting.
• Prompt notes: "Same instructor avatar every episode (seeded by context images), blackboard-style callouts, clean captions, sections recap."

• YouTube long-form + shorts: Generate a main video and a 60s cut.
• Ecommerce: Embed the product story on PDPs and email post-purchase content.
• Social: Auto-post reels to Instagram/TikTok with a second n8n branch.
• Internal training: Archive to a knowledge base and share to Slack.

• API & VDK: Build programmatic pipelines to generate and customize videos at scale, then let n8n orchestrate triggers and distribution. Explore Scrptly’s developer platform via https://scrptly.com/
• MCP server: Connect Scrptly with your preferred LLM through MCP to let your agent reason about prompts, research, and scene planning before generation.

• Store your Scrptly API key in n8n Credentials, not hard-coded.
• Protect context images URLs; prefer secure storage and short-lived links.
• Log which prompts and assets were used per output video for provenance.

• Video feels disjointed? Add explicit scene beats and transitions to the prompt.
• Characters or products look inconsistent? Provide multiple high-quality context images (front/side/detail shots) and reuse them.
• Timing off? Specify target durations per segment and total length.
• Long jobs timing out? Switch to async mode and poll with retries.

• Install the Scrptly node in n8n and connect your API key.
• Start with the long-form prompt template above, attach a few strong context images, and run a test.
• Scale by adding async generation, error handling, and multi-channel publishing.