Text-to-Video
Write the video prompt first, then use model, duration, aspect ratio, and resolution settings to generate ad-ready short videos.
Text-to-Video is designed for script-first video generation. The page combines Prompt, Negative Prompt, model choice, aspect ratio, resolution, duration, and output count, then saves generated videos into a reusable gallery.
When To Use
Use this page when the message direction is already clear, or when Storyboard has already given you the right shot logic.
It is effective for ad-video drafts, pacing tests, and multi-version experimentation around the same selling point.
Workflow
The most reliable pattern is prompt first, parameters second, and gallery review third.
- 1
Step 1: Write the Prompt and Negative Prompt
The main prompt defines subject, scene, and motion. Negative Prompt helps suppress unwanted artifacts, distortion, or watermark-like output.
- 2
Step 2: Set Model, Aspect Ratio, Resolution, Duration, and Count
These parameters shape pacing, channel fit, and how many versions you generate in one batch. Start small while the direction is still unproven.
- 3
Step 3: Manage Results in the Video Gallery
Completed tasks appear in the gallery, where you can refresh status, download videos, edit again, or batch-delete.
Key Controls
The biggest quality differences usually come from the prompt, negative prompt, and duration choices.
- Prompt: defines the main narrative and shot language
- Negative Prompt: removes recurring visual problems
- Model: balances speed and generation quality
- Resolution / Aspect Ratio / Duration: decides channel fit and pacing
Best Practices
Start with shorter clips and fewer outputs while you validate direction, then scale the best prompt into higher resolution or more versions.
Use Edit Again from the gallery for small improvements instead of rewriting the entire request every time.
- Prompt should describe value proposition first, then motion and framing
- Negative Prompt should target your most common issues, not become a random word dump
- Download and organize winning outputs quickly so they can become reusable assets