verified Content Guidelines

Content Guidelines

Learn how to write prompts that comply with AI model moderation policies

info

Important Notice

VideoGenAPI does not apply its own content moderation. All moderation is performed by the AI model providers themselves (OpenAI, Kling, Google, etc.). Each model has its own content policies that we cannot override.

policy

1. Understanding Content Moderation

Content moderation is implemented by the AI model developers to ensure responsible use:

smart_toy Model-Level Moderation: Each AI model (Sora, Veo 3, Kling, etc.) has built-in safety filters
shield Automated Detection: Prompts are analyzed before video generation begins
block Rejection Response: Flagged prompts return an error without generating video
block

2. Prohibited Content Categories

The following content types are generally prohibited by AI model providers:

dangerous Violence & Gore: Explicit violence, blood, injuries, or graphic content
warning Adult Content: Nudity, sexual content, or explicit material
sentiment_very_dissatisfied Hate Speech: Discriminatory, racist, or hateful content
local_police Illegal Activities: Drug use, illegal weapons, or criminal acts
coronavirus Harmful Content: Self-harm, suicide, or dangerous challenges
child_care Child Safety: Any content involving minors in unsafe situations
check_circle

3. Best Practices for Writing Prompts

Follow these guidelines to create prompts that pass moderation:

light_mode Be Positive & Creative: Focus on constructive and imaginative scenarios
public Use Neutral Language: Avoid loaded terms or controversial subjects
visibility Be Specific & Clear: Detailed prompts help avoid misinterpretation
family_restroom Keep It Family-Friendly: Assume your content should be suitable for all ages
lightbulb

4. Examples of Acceptable Prompts

Here are examples of prompts that typically pass moderation:

check_circle

Cinematic & Artistic

"A cinematic drone shot flying over a misty mountain range at sunrise, golden light rays breaking through clouds"

check_circle

Nature & Wildlife

"A playful dolphin jumping out of crystal clear ocean water in slow motion, with water droplets sparkling in sunlight"

check_circle

Urban & Architecture

"Time-lapse of a busy Tokyo street intersection at night, neon lights reflecting on wet pavement"

check_circle

Fantasy & Sci-Fi

"A magical library with floating books, glowing ancient manuscripts, and sparkling dust particles in volumetric light"

cancel

5. Examples of Problematic Prompts

These types of prompts are likely to be rejected:

cancel

Explicit Violence

Prompts describing injuries, blood, fighting, or weapons in detail

cancel

Inappropriate Content

References to nudity, sexual acts, or adult situations

cancel

Discriminatory Language

Content targeting protected groups or promoting stereotypes

cancel

Dangerous Activities

Illegal substances, criminal behavior, or harmful acts

edit

6. How to Modify Rejected Prompts

If your prompt is rejected, try these strategies:

swap_horiz Use Alternative Words: Replace sensitive terms with neutral alternatives
zoom_out Increase Distance: Make scenes more abstract or distant
palette Focus on Aesthetics: Emphasize artistic elements over specific actions
context Add Context: Clarify the purpose is educational, historical, or artistic
psychology

7. Model-Specific Considerations

Different AI models may have varying levels of strictness:

star Sora (OpenAI): Strict moderation following OpenAI's usage policies
star Veo 3 (Google): Comprehensive safety filters for brand protection
star Kling: Moderate approach with focus on violent/sexual content
info Tip: If one model rejects your prompt, try another model
troubleshoot

8. Handling Moderation Errors

When you receive a content moderation error:

search

1. Review Your Prompt

Identify potentially sensitive words or concepts

edit

2. Rephrase & Simplify

Use clearer, more neutral language

refresh

3. Try Again

Submit your revised prompt

support_agent

4. Contact Support

If you believe your prompt was incorrectly flagged

contact_support

9. Need Help?

If you need assistance with content moderation issues:

email Email: contact@lannetech.com
description Include your prompt and the error message received
schedule We typically respond within 24 hours

Ready to create amazing videos?

Follow these guidelines and start generating stunning AI videos

Get Started arrow_forward View Documentation description