How To Start Making AI Videos in 2026 (Beginner Guide)
This guide provides working AI filmmakers with a credit-conscious reference workflow, detailing how structured multi-panel character sheets and environment references prevent costly generation failures across video models.
What this lesson covers
Dan Kieft outlines an asset-first framework for beginner AI video creation, emphasizing consistent reference generation prior to rendering video clips. He demonstrates generating multi-panel character sheets and environment references using Higgsfield's Cinema Studio and LLM prompting templates.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 01:43 ↗
Choosing a platform that aggregates multiple AI video models prevents lock-in to single models that quickly become outdated.
- 02:43 ↗
Using reference assets (character, location, prop sheets) yields far more visual control and consistency than direct text-to-video generation.
- 03:20 ↗
A structured AI video pipeline requires completing ideation and asset generation before moving into video rendering and assembly.
- 04:39 ↗
Image generation is significantly cheaper than video generation, making iterative image asset refinement the primary strategy for preserving credits.
- 05:18 ↗
Structured LLM prompt templates can reliably format multi-panel reference descriptions showing full-body front, back, and close-up views of a character.
Workflow outlined in the video
- Access Higgsfield's Cinema Studio and navigate to the AI Cast feature under the image generation tab.
- Generate a character sheet using detailed physical descriptors (age, ethnicity, hairstyle, outfit, heels) at 2K resolution.
- Paste a three-panel character reference formula into an LLM to generate expanded, structured prompts for custom characters or selfie references.
- Test generated reference prompts across different image models (such as Cream 5.0 Pro or Nabana) in 16:9 aspect ratio.
- Tag reference images (e.g., '@image1') in video generation model inputs to lock character visual attributes.
- Create location reference sheets within Cinema Studio's cinematic locations tool to maintain recurring environment consistency.
Before you use this workflow
These notes describe the source video at its publication date. Model access, pricing, connectors and interfaces may have changed. Check the original source and the provider’s current documentation before installing an add-on, connecting an account or spending credits.
ReelStack has not independently tested this workflow. Preview one representative shot and check motion, continuity and output quality before applying it to a full production. No result, cost saving or model capability is guaranteed.
Explore the tools
How to interpret the numbers
107,583 views captured 30 October 2026. ReelStack recommendation score: 50/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 65% performance: 0.99× views vs 8 other channel videos observed at a similar age. View-evidence factor 99% (views / (views + 1,000))
- 25% freshness: 68/100 with a 14-day half-life
- Momentum pending: collecting daily snapshots; its 20% weight goes to observed performance, not free points
- 10% engagement: 4/100; likes + 4× comments, smoothed with a 500-view neutral prior