> ## Documentation Index
> Fetch the complete documentation index at: https://docs.bizyair.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Quick Create

> Generate images, videos, and audio right in your browser — no coding required. Supports text-to-image, image-to-video, and more.

Quick Create is a lightweight, one-click creation tool offering image and video modalities, each with a dedicated model pool. It supports text-to-image, image-to-image, text-to-video, and image-to-video — integrating prompt input, reference upload, aspect ratio / quality / duration settings, compute management, and automatic language translation into a single streamlined workflow for fast, high-quality output.

## Modalities and Model Selection

| Modality | Supported Generation Modes     | Model Pool          |
| -------- | ------------------------------ | ------------------- |
| Image    | Text-to-Image / Image-to-Image | Image model library |
| Video    | Text-to-Video / Image-to-Video | Video model library |

<Note>
  Model pools and parameter configurations are isolated by modality. Switching modalities will reset your current selection.
</Note>

## Reference Upload

* Supports uploading a single reference image or video, labeled as "Reference."
* All uploaded media goes through platform safety screening — violating content is automatically blocked.

## Prompt Input

* Enter a positive prompt describing your target image, shot, or content.
* Overly long text triggers a Prompt Validation Error; sensitive content triggers a safety block.
* Multi-language input is supported — the platform automatically translates to the optimal language for the target model.

## Generation Parameters

Available parameters vary slightly by modality. The following are common parameters:

| Parameter         | Type    | Default         | Description                                     |
| ----------------- | ------- | --------------- | ----------------------------------------------- |
| `prompt`          | string  | Required        | Positive prompt describing the target content   |
| `negative_prompt` | string  | —               | Negative prompt (optional)                      |
| `width`           | integer | 1024            | Output width in pixels                          |
| `height`          | integer | 1024            | Output height in pixels                         |
| `steps`           | integer | Model-dependent | Number of sampling steps                        |
| `guidance`        | number  | Model-dependent | Guidance strength                               |
| `seed`            | integer | -1              | Random seed; use -1 for random                  |
| `duration`        | number  | Model-dependent | Video duration in seconds (video modality only) |
| `output_format`   | string  | png/jpeg        | Output format                                   |
| `num_images`      | integer | 1               | Number of images per generation                 |

## Best Practices

1. For image-to-image, upload a reference image with a composition close to your target — this significantly reduces the need for prompt engineering.
2. Fix `seed` for reproducible results; increase `temperature` or change `seed` for more variety.
3. Video generation time scales with duration — start with a short clip to validate before scaling up.
4. Describe style, lighting, composition, camera, and details in your prompt rather than using generic descriptions.