Guide to Video Generation Parameters

What Are Video Generation Parameters

Before generating a video, you can adjust a few basic settings that define how your output looks and behaves:

• Aspect ratio (the width-to-height shape of your video)
• Length
• Number of variations
• Native sound generation.

These options are all available right in the prompt box at the bottom of the tool.

🎥 Auto Mode

Auto Mode is the default and what we recommend for most users. This way, the right video generation model is chosen automatically based on your prompt.

Available options in Auto Mode

• Aspect ratios: 16:9, 9:16 (landscape or vertical)
• Lengths: 5, 6, 8, 10, 12 seconds (without sound) / 8 seconds (with sound)
• Native sound: on or off.
• Variations: 1, 2, or 4
• References: supports First Frame, First + Last Frame, or multiple image References. You can't use both First (or First + Last) Frame and References simultaneously.

Auto Mode also includes smart prompt enhancement, which expands short prompts into detailed, well-structured descriptions for smoother, more consistent results.

⚙️ Custom Mode

If you want to explore or test different setups, you can switch to Custom Mode for more control.

To do this, click the settings icon (next to the video count selector) and set the toggle to Custom.

In this mode, you can:

• Choose a model manually
• Pick from different clip lengths and aspect ratios (varies by model)
• Toggle Prompt Enhancement on or off.

Reference options in Custom Mode

Each model handles references differently:

• Some allow First Frame only
• Others allow First + Last Frame
• Some support multiple reference images.

You’ll see all supported features listed right under the model name once you select it.

That’s it: set your ratio and duration, add any references you want, and generate!