How to Make Your First AI Video with Higgsfield AI!
This video offers a practical breakdown of multi-shot text-to-video prompting structures and pre-production image workflows for maintaining character and asset consistency in AI video generation.
What this lesson covers
Youri van Hofwegen demonstrates how to generate cinematic AI videos on Higgsfield AI using models like Seedance 2.5, Juan 3.0 Prime, and MiniMax H3. He walks through writing multi-shot structured prompts and generating character reference sheets to maintain visual continuity across video scenes.
Key takeaways from the creator
AI-extracted notes, not independently verified product claims. Timestamp links let you check each point in the original video.
- 01:08 ↗
Seedance 2.5 supports generating full 30-second realistic video clips with dialogue from a single multi-shot prompt, with a draft mode at 480p available to test prompts before spending credits.
- 03:12 ↗
Juan 3.0 Prime excels at stylized anime and fast-action scenes while maintaining visual coherence across rapid movement and explosions.
- 04:23 ↗
MiniMax H3 offers a cost-effective option for rendering video generations up to 2K resolution for shorter 15-second clips.
- 05:27 ↗
Achieving consistent characters, props, and locations across multiple video shots requires creating foundational base images before feeding them into video generation models.
- 07:00 ↗
Character sheets with three angles (front, back, and facial close-up) on a flat gray background prevent facial morphing and unwanted lighting baked into references.
Workflow outlined in the video
- Select the appropriate video model based on project needs, using Seedance 2.5 for realism and dialogue, Juan 3.0 Prime for stylized anime action, or MiniMax H3 for 2K resolution.
- Structure multi-shot video prompts starting with style and context, followed by individual shot breakdowns describing specific camera movement, action, dialogue in quotes, and sound effects.
- Use draft generation modes (like Seedance 2.5 draft at 480p) to test shot sequencing and camera motion before committing credits to final renders.
- Generate baseline asset images (logos, environments, and product bottles) using consistent style descriptions and negative prompts to eliminate unwanted visual elements.
- Build reference character sheets containing full-body front/back views and close-up face shots on neutral gray backgrounds to maintain facial continuity across video clips.
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
38,081 views captured 12 November 2026. ReelStack recommendation score: 63/100. Formula: radar-v5. This is ReelStack’s calculation, not a YouTube rating or a measure of factual accuracy.
Show the calculation
- 45% performance: 1.59× views vs 33 other channel videos observed at a similar age. View-evidence factor 97% (views / (views + 1,000))
- 25% freshness: 87/100 with a 14-day half-life
- 20% momentum: 6799 views/day vs 2933 channel baseline (33 comparable uploads), with the same view-evidence factor
- 10% engagement: 2/100; likes + 4× comments, smoothed with a 500-view neutral prior