Create Images

Feature 1: Generate image based on your daily language. (Text to Image)

Feature 2: Generate image based on uploaded image and the way you want to change it. (Image + Text to Image)

Text to Image

Set parameters in the left column. (If not intentionally set, use the default parameters when generating)

Model: choose the style you like for the image. Each model leads to distinct style and capability. We have many models, with two core models（strongly recommened）: Real and Anime, and several featured models, including Ink Painting, Pixar, Ghibli, Mosaic, and more. EXPLORE AND DISCOVER!
1. Real: Suitable for all real world images.
2. Anime: Can be used for all animation.

For the featured model, you can adjust the weight. This will decide how strongly your chosen model is weighed. Smaller means welcome more randomness, and bigger means wish to stick to the style of the mode. Adjust this number carefully, our default is 0.65. If this number is too small, your chosen model may not be valued in the output. If this number is too big, it may introduce too many noises and leads to a bad outcome as well. So ADJUST CAREFULLY!

Image Dimensions: choose aspect ratio of the image.
Number of Images: choose the number of images you want to generate this time, remember more images will cost more tokens. (2 tokens/ image)
Text Guidance Scale: How strongly your prompt is weighed. Similar to the weight of the chosen model, adjust this number carefully as well.
[Upgrade Plan] No Watermark: No watermark on the output.
[Upgrade Plan] Private Mode: Keep your generations private, invisible to other userd and LensGo AI.
Negative Prompts: Type in want you don't want to see in the image. This could be anything: "low quality", "three legs", "horror", "text", etc.

Note:

When hovering on any question mark, you'll see the explanation for that setting. So if you are uncertain about the settings, just hover over the question mark next to the parameter.

Default paramters are set to be: Model(Real), Image Dimentions(1:1), Number of images(2), Text Guidance Scale(7), No Watermark(No), Private Mode(No), Negative Prompts(No)