Image Generation

Overview

NovaAI Image Generation uses advanced text-to-image models to produce images from descriptive prompts. Simply describe what you want to create, configure the generation parameters, and the model returns one or more generated images.

The quality of the output depends on the clarity and specificity of your prompt. Providing visual details such as style, composition, lighting, colors, and camera perspective helps the model generate more accurate results.

How Image Generation Works

Generate Your First Image

1import { NovaAI } from "@novaai/sdk";
2 
3const client = new NovaAI({
4  apiKey: process.env.NOVA_API_KEY,
5});
6 
7const image = await client.images.generate({
8  model: "nova-image-1",
9  prompt: "A futuristic smart city at sunset with flying vehicles, cinematic lighting, ultra detailed.",
10  size: "1024x1024"
11});
12 
13console.log(image.data);

Request Properties

Image generation model.

Text description of the desired image.

Output image dimensions.

Image quality level.

Image output format.

Number of images to generate.

Prompt Writing Tips

High-quality prompts typically include:

• Subject

• Environment

• Artistic style

• Lighting

• Camera perspective

• Color palette

• Composition

• Level of detail

Example Prompt

The more descriptive your prompt, the more predictable the generated image becomes.

Supported Image Sizes

Response Example

{
  "id": "img_91af7d",
  "created": 1743256821,
  "data": [
    {
      "url": "https://cdn.novaai.dev/images/image-01.png"
    }
  ]
}

Best Practices

• Write clear and descriptive prompts.

• Include visual style references when appropriate.

• Specify lighting, composition, and camera angle.

• Choose the correct output resolution for your use case.

• Generate multiple variations when exploring creative concepts.

Common Prompt Examples

Product Photography

Illustration

Concept Art

Common Errors

Empty Prompt

The request does not contain a valid prompt.

Solution

Provide a descriptive text prompt before sending the request.

Unsupported Resolution

The selected image size is not supported by the model.

Solution

Choose one of the officially supported output dimensions.

Generation Timeout

The request exceeded the maximum processing time.

Solution

Retry the request or reduce the number of requested images.

Frequently Asked Questions

Next Step

You've learned how to generate images from natural language prompts. Continue to Embeddings to discover how NovaAI converts text into vector representations for semantic search, recommendations, and retrieval-augmented generation (RAG).