Image Generation

RenderAI allows users to create stunning images using generative AI, via four different methods.

Image Generation Modes

Text-to-Image

The user describes the image they want in text.

Image-to-Image

The user uploads an image for the AI to use as inspiration, and (optionally) describes the output further with text.

Sketch-to-Image

The user draws a rought sketch for the AI to use as inspiration.

Inpainting

The user uploads an image and highlights only a portion of it, then describes the desired changes that the AI will make within the selected area.

Prompt Engineering

Generative AI works on word association. That means, to get best results, you will need to use words that will trigger a response, even if they are over the top, or absurd. It is also important to describe the desired medium.

For example: "a dog wearing a suit" may result in a cartoon, but "a photograph of a dog wearing a suit, shot on film, 50mm lens, bokeh, award winning photography" will produce a photorealistic result.

Settings

Negative Prompt

Aanything you don’t want to see.

Width & Height

The dimensions of the image.

Inference Steps

The higher the number of steps, the higher the quality, but more steps take more time.

Randomize Results

If you want a random seed each time, leave this checked.

Seed

If you want to use a specifcic seed, set it here.