StoryTool logo
Back to Knowledge Base

Creating Videos

From Script to Screen: The Complete Guide to Generating Your First Video.

1. The 6-Step Workflow

StoryTool is designed for speed. Create a ready-to-publish video in just 6 steps:

1

Paste your text

Input your script (supports up to ~60k characters).

2

Choose Style & Voice

Select from 26 visual styles and 20+ languages.

3

Select Agent & Ratio

Choose Story Agent (Fiction) or Edu Agent (Info) and the aspect ratio (16:9 or 9:16).

4

Add Branding

Upload your Intro, Outro, and select background music.

5

Metadata (Optional)

Click "Gen Title" to let AI suggest SEO titles and descriptions.

6

Generate

Click the button to start rendering.

2. Best Practices for High-Quality Results

A. Input Standardization (Crucial)

Garbage in, garbage out. The AI needs clean data to perform well.

  • Logic & Spelling: Check for spelling errors and logical flow. Misspelled names lead to inconsistent character visuals.
  • Punctuation: Use proper commas and periods. Long, run-on sentences degrade audio quality.
  • Technical Terms: Avoid obscure abbreviations.

Pro Tip: Use an LLM (like GPT/Gemini) to "standardize text for TTS" before pasting it here.

Test Audio First: Use the Creating Audio Page to generate a sample voiceover. If the audio is bad, the video will be bad.

B. Art Direction & Context

  • Match the Vibe: For Informational content, use the Edu Agent with a fresh, medium-fast voice. For Stories, use the Story Agent.
  • Define the Context: AI guesses based on text.
    • Vague Input: "I walked down the street." → AI guesses a random generic street.
    • Clear Input: "Paris, 1850s. I walked down the cobblestone street." → AI generates 19th-century French architecture.
    • Note: If your story is in French, the AI defaults to a French setting unless specified otherwise.

3. The Process & Output

Execution

  • You must Verify Twice before generation.
  • You can close the browser/page after clicking Generate. The server handles the rendering.
  • Wait Time: Results can take from 8 minutes up to 4-8 hours depending on input length.

What You Receive (Project Assets)

At the top of your Project Page, you will find:

Images Zip

High-res source slides

Audio Zip

Isolated voice tracks

Video File

HD 1080p, 60 fps

Final render (with/without subtitles)

SRT File

Soft subtitles for multi-language dubs

Metadata File: Title, description, hashtags included.

Tip: Use the SRT file in the "Gen Audio" page to create Multi-language Dubs for YouTube.

Disclaimer & Safety

  • User Responsibility: You are liable for copyright and community guideline compliance.
  • Prohibited Content: Requests containing heavy violence, gore, NSFW, or inappropriate child-related content will be blocked or fail. No refunds for policy violations.
  • Review: You agree to review all AI-generated content before publishing.

4. Understanding AI Limitations(State of 2025)

Honesty is our policy. Even SOTA models (GPT, Gemini, Claude) have limits.

Visual Consistency

  • Error Rate: Low, but possible. You might see small inconsistencies in character hair, clothing details, or background objects changing between frames.
  • Historical Accuracy: AI struggles with micro-details (e.g., specific military rank insignias, exact tools from a niche historical period).
  • Spatial Logic: 3D spacing and relative distances may occasionally be imperfect.

Solution: We accept these current AI limitations (as of Dec 2025). For perfection, use the provided Asset Pack to manually edit/replace images using external tools (Nano Banana/Qwen Edit) and re-assemble in CapCut.

Voice Cloning

Always test cloned voices in the Audio Page first. Bad clones waste video credits.

System Overload

Rarely, high traffic may cause errors. Report unusual failures to info@storytool.io for credit refund assessment.

5. Support

Need help? Contact us at: info@storytool.io (Typical response within 24h).

Make Good Myth and monetize your creativity! :D