TotalApp Docs

Text to Video

Generate dynamic video clips from text descriptions using TotalApp's advanced AI diffusion pipeline — no camera, footage, or editing experience required.

Overview

Text to Video is TotalApp's core AI video generation tool. Type a natural language description of a scene and the AI renders a smooth, temporally consistent video clip matching your description. The engine supports multiple artistic styles — from photorealistic to anime — and gives you fine-grained control over motion speed, output quality, and clip duration.

Under the hood, Text to Video uses a latent diffusion model with a temporal attention module that ensures frame-to-frame coherence, preventing the flickering and incoherence that characterise lower-quality video AI tools.

Natural Language Input

Describe any scene in plain English. No special syntax or keywords required — the model interprets descriptive prose, adjectives, camera directions, and mood naturally.

Five Artistic Styles

Choose Anime, 3D Animation, Clay, Comic, or Cyberpunk. Each style applies a distinct visual grammar to the generated footage while preserving your scene description.

Negative Prompt Support

Specify what to exclude from the output — unwanted elements, artefacts, or colour palettes — for cleaner, more focused results every time.

Reproducible Outputs

Pin a seed value to reproduce the exact same video when refining prompts or sharing settings with collaborators.

Input Parameters

Every parameter is accessible from the Text to Video panel. Configure them before generation; most can be changed and regenerated without additional charge until you download.

Parameter Options Description
Text Prompt Free text, up to 500 characters Primary description of the scene, subject, action, camera movement, and mood
Negative Prompt Free text, up to 500 characters Elements to exclude — e.g., "blurry, low quality, watermark, text overlay"
Style Anime, 3D Animation, Clay, Comic, Cyberpunk Overall visual aesthetic applied to the generated frames
Duration 5 s, 8 s Length of the output clip; 8 s is better for complex or multi-element scenes
Quality 360p, 540p, 720p, 1080p Output resolution; higher quality increases generation time and credit cost
Motion Mode Normal, Fast Controls motion intensity — Fast produces more dramatic, energetic movement
Seed Random, Fixed integer Fix to reproduce identical outputs; leave random for diverse generations

Output Specifications

Specification Value
Output formatMP4 (H.264), silent
Frame rate24 fps
Maximum resolution1920 × 1080 (1080p)
Average generation time~45 seconds
Max prompt length500 characters per field

How to Generate a Video

Open Text to Video
Write Prompt
Select Style
Set Parameters
Generate
Download
  1. Navigate to Video Creator → Text to Video in the TotalApp sidebar.
  2. Enter your scene description in the Text Prompt field. Be specific: include subject, action, camera movement, lighting, and mood.
  3. Optionally add a Negative Prompt listing elements to exclude.
  4. Select an artistic style from the five available options.
  5. Choose Duration (5 s or 8 s), Quality, and Motion Mode.
  6. Click Generate. Generation typically completes in ~45 seconds.
  7. Preview the result. If satisfied, click Download to save the MP4.

Prompt Writing Guide

Anatomy of a Strong Prompt

A well-structured prompt follows this pattern: [Subject] [Action] [Environment] [Camera] [Lighting] [Style cue]

Example: "A lone astronaut floating weightlessly through the International Space Station corridor, slow-motion drift, blue-white LED ambient lighting, cinematic wide shot"

Style-Specific Prompt Tips

  • Anime: Reference anime conventions — "sakura petals falling", "dramatic wind effect", "speed lines"
  • 3D Animation: Use terms like "Pixar-style", "subsurface scattering", "volumetric lighting"
  • Cyberpunk: Mention "neon reflections", "rain-slicked streets", "holographic displays"
  • Clay: Describe subjects as "clay-textured", reference "stop-motion feel", "muted pastel colours"
  • Comic: Use "bold outlines", "flat colour", "halftone dots", "panel framing"

Effective Negative Prompts

Common negative prompt elements to include for cleaner output: blurry, low quality, watermark, text overlay, extra limbs, deformed hands, face distortion, flickering, static noise

Frequently Asked Questions

Why does the AI sometimes misinterpret my prompt?
The model interprets prompts holistically. Ambiguous descriptions — "a good scene", "something interesting" — give the AI too much latitude. Add specific adjectives, a clear subject, and an explicit action. If results are still unexpected, simplify the prompt first and add detail incrementally.
Can I generate the same video twice?
Yes. Set a fixed seed value before generation. The same prompt + same parameters + same seed will always produce the same output. Note that if any parameter changes, the output changes too even with the same seed.
How do I get smoother motion?
Use Normal motion mode and describe slow or deliberate movements in the prompt ("slowly", "gentle", "drifting"). Fast mode amplifies motion which can produce jarring results on subtle scenes. Choose 8 s duration for more time for the motion to develop naturally.
Is 360p output usable for social media?
360p is adequate for quick previews and experimentation but appears noticeably soft on modern phone screens. Use 720p as the minimum for published social media content and 1080p for anything shown on a desktop screen or included in presentations.