Dream Building Industry | Niji Vision XL

Hope is the only thing stronger than fear, a little hope can bring a person back to life — Donald Sutherland (2013)

V2.0 - Tranquil Sea

Developer Update Notes:

Welcome to Niji Vision, this time we welcome a major update of the V2.0 version!V2.0 is an iterative version based on the training of V1.0. If you have the previous version of this model, it is recommended to update to V2.0 immediately for a better drawing experience!(Important!)When using V2.0 for drawing, continue using the Stylize Parameter from V1.0, for details please refer to the V1.0 model instructions.V2.0 has added more materials on top of 1.0, including materials with different artistic styles, allowing the model to have a wider range of drawing styles. For example, Sketch Art, Pixel Art, etc., have been greatly enhanced in V2.0. V2.0 has better limb stability and more excellent performance in multiplayer interaction scenes. It also has better color understanding feedback, V2.0 can use more complex lighting cues to control the image more accurately. Additionally, it includes enhanced training on more sci-fi materials, making the sci-fi effects of V2.0 even cooler.All training sets of this model have been processed withWellCap for natural language labeling. For image output, it is recommended to use standard natural language descriptions following NLP guidelines to describe the scene you want. The model's prompt feedback is very high, making it possible to generate images based on the input. It is recommended to have a base image resolution of no less than 1024 x 1024 for the ultimate visual experience. For online rendering, it is recommended to enable high-definition zoom for better visual effects. For more examples, you can refer to my model showcase.(For detailed examples, please refer to my model showcase, 20 images in total, highly recommended to view all for a better understanding of this model! Thank you!)

Warning:

According to Midjourney TOS Article 4, subscribing users have the copyright to all output images of Midjourney & Niji. To avoid unnecessary misunderstandings, please do not upload this model to any other model platform, do not use images produced by this model to train new models, and prohibit all forms of model fusion. Thank you!