Models
Last updated
Last updated
When you’re in the Video Generation tab, you’ll see a list of models—each with unique strengths and ideal use cases. Here's a breakdown of each model and how to get the best results based on your prompt.
Standard Video Generation Model
Best for: General video generation, standard quality, and flexibility for most prompts.
Prompting Tips:
This model is a great all-rounder for both simple and complex scenes.
Use descriptive language with clear subject-action pairs:
“A robot walks through a futuristic city street, its glowing eyes scanning the area.”
For general settings, avoid excessive camera direction unless specified.
More Camera Control for Detailed Shots
Best for: Users who need precise camera direction or more cinematic control over shots.
Prompting Tips:
Explicitly describe camera movements such as “zoom in,” “pan right,” or “track forward.”
Example prompt:
“Close-up of a robot’s face, zoom in to its glowing eyes, then pan left to reveal the city skyline in the background.”
Perfect for dynamic scenes where you want the camera to follow the action closely.
Advanced Model with Enhanced Quality
Best for: High-quality, complex scenes with detailed textures and background elements.
Prompting Tips:
Ideal for scenes with multiple layers of action or backgrounds (e.g., crowds, detailed environments).
Use clear separation of elements:
“A bustling futuristic city with flying cars zooming overhead, neon signs flickering in the background, and people walking on busy streets.”
Add camera movements to bring focus to key elements in a busy scene.
Balanced High-Quality Model for Complex Scenes
Best for: Scenes requiring a balance of realism and artistic interpretation—especially when you need a mix of dynamic environments and rich textures.
Prompting Tips:
Best for detailed environments and characters with complex features.
Specify the mood and setting to get the most out of this model:
“A battle between two samurais in a misty forest, their swords clashing as the camera circles them.”
Try adding lighting and atmosphere cues to enhance the depth of the scene.
Slow, High-Quality Model with Good Prompt Understanding
Best for: Detailed, realistic outputs that focus on character development or intricate environments.
Prompting Tips:
Ideal for slower, dramatic scenes or highly detailed character shots.
Keep prompts clear and concise:
“A lone knight stands in a misty field, his armor reflecting the moonlight as he gazes at a distant mountain.”
Use slow-moving actions or static scenes to allow the model to focus on high-level details.
Aspect Ratio: Pick 9:16 for vertical, 1:1 for square, or 16:9 for cinematic landscape.
Action and Setting: Always start by clearly describing what’s happening and where it’s happening.
Lighting and Camera Work: Models like Hailuo Director and WAN 2.1 benefit from specific camera movements and lighting cues.
Iterate and Experiment: If a scene doesn’t look right, adjust the camera work or action descriptions and try again.
With these models and tips, you’re set to create immersive, dynamic video clips with ease. Let your imagination run wild and see how these tools bring your vision to life!
